Understanding and Changing the World

Joseph Sifakis

Understanding and Changing

the World
From Information to Knowledge
and Intelligence
Joseph Sifakis
Verimag Laboratory
Université Grenoble Alpes
St Martin d’Hères, France

ISBN 978-981-19-1931-2 ISBN 978-981-19-1932-9 (eBook)

https://doi.org/10.1007/978-981-19-1932-9

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore
Pte Ltd. 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Πάντες ἄνθρωπoι τoῦ εἰ δεναι ὀρεγoνται
φύσει—All men naturally desire knowledge.
Aristotle, Metaphysics
Preface

I would like to briefly discuss the motives that have prompted me to write a book
with such a broad and ambitious objective that goes beyond the scope of my specific
scientific and technical knowledge.
In the early 1980s, I came to realize that my research into applied logic and
mathematical language theories was directly related to philosophical problems,
particularly the problem of consciousness and language.
I thus started reading and became interested in philosophy. I must say that, despite
my best efforts, I could not find a path through the bewildering maze of philosoph-
ical theories. It bothered me that many philosophers are not strict enough in their
thinking. Their discourse is nearer to the literary. This may well be due to certain
individuals who happen to be fine wordsmiths, who know how to write elegant and
exciting text. However, most philosophers did not convince me, as I am possessed of
the strict, pedestrian mind of an engineer. I find that many play around with
terminology, creating new terms without actually caring about how they relate to
existing ones. What has particularly struck me is the existence of closed communi-
ties around a philosophical or even scientific trend. Organized by a clergy of
“authorities,” they use hermetic terminologies that render their theories inaccessible
to the “uninitiated” and lends them legitimacy in the eyes of laymen. Thus, there are
existentialists, structuralists, empiricists, Marxists, Hegelians, etc., each philosophiz-
ing from their own context and pulpit, clinging to their “truths” and disregarding
everyone else, without a thought to how their knowledge helps people live a life
worth living. This may explain why the present day is characterized by a depreci-
ation of philosophy and the humanities, by a dearth of philosophical education.
I believe that any attempt to formulate our knowledge must respect the laws of an
economy of thought, must be based on a set of well-founded concepts that enjoy
clearly defined semantic relationships with each other.
After much philosophical perambulation and searching, I decided in the late
1990s to strictly formulate my views on the world starting with what I am sure
and aware of. The truth, as it is created and processed by the human mind, cannot by
its very nature be complex. All theories and knowledge of any utility are shockingly

vii
viii Preface

simple in their principles—regardless of whether we use specialized and sophisti-

cated techniques to study them. I believe that it is dangerous for our societies to
become alienated from scientific and technical knowledge. We must give them the
tools to judge, at the very least, its importance and the impact of its application.
I wanted to gradually build an outlook of the world where I always distinguish
what I can talk about with some degree of certainty from what I do not know. Our
aim must be to push the boundaries of knowledge as far as possible, while remaining
aware of our limitations.
In my searching, I quickly identified the issues of knowledge and, therefore,
consciousness and language as the key themes.
For decades, I have been grappling with human intelligence, whether it is possible
for a computer to approximate it. Can one study human behavior—while avoiding
the traps of simplification—and draw certain useful, practical conclusions, for
example, on what freedom of will is and what its management means, what a
deadlock and what its solution consists of?
I reached the conclusion that the world is the combination of two distinct, yet
complementary entities: matter/energy on the one hand and information on the
other—what can be seen and what can be thought.
I despise moralism and vagueness. I believe that even ethical problems have a
technical dimension that we must understand in order to live better lives having,
naturally, first identified the limits of such an approach.
What has prompted my investigations and inquiries is the realization that we are
living in a time of absolute relativism and confusion. The crisis we have been living
through for decades has gravely necessitated a philosophy of the individual. One that
explains in a simple and comprehensible manner, free of dogmatism, the practical
necessity of values and ethics for individual happiness and social peace and pro-
gress. One that helps us understand the importance and limits of knowledge, deepen
our self-knowledge, and delve into the dual-nature, double-faceted game of com-
prehension: conscious and subconscious, logical and intuitive, empirical and
transcendental.
It is shocking how real problems are being thoroughly hidden from the public and
how the conditions for concealing them are artificially created. At the same time,
these real problems are replaced by other fictitious ones, which disorient people and
divert them from the conscious search for solutions. Scientific and technical progress
is useless, if not dangerous, for societies unworthy of managing and directing
it. Even the most perfect legal system is ineffectual without free compliance with
ethical rules. Lack of intellectual values cannot be offset by tangible goods.
Over the course of my research and academic career spanning 50 years, my work
has focused on building cost-effective, trustworthy computerized systems. My early
work on systems verification met with broad application in systems engineering. I
later studied system design as a process that leads from technical requirements to
artifacts satisfying them. This led me to investigate the role of knowledge in
problem-solving and its inherent complexity limitations. Over the past 10 years, I
have worked on the design of autonomous systems and of self-driving cars in
Preface ix

particular. My research on autonomous systems design has deeply informed and

inspired the ideas about intelligence and consciousness developed in the book.
I believe that the criterion for intelligence is not just a Q&A game. It is about
building systems that can replace human agents performing tasks in complex
organizations. Therefore, apart from solving hard computational problems, it also
involves challenging systems engineering issues that have been often
underestimated.
It was my wish and responsibility to write about the importance of knowledge as
an intangible entity, separate from physical entities, and the need to understand its
development and implementation processes; to show how new concepts and models
stemming from informatics can help us deepen the functioning of individual and
collective consciousness; and to demonstrate how all the structures that make up our
societies implement above all else information systems whose proper functioning
depends on the seamless and timely exchange of reliable knowledge.
It is my hope that this endeavor marks the first step in the reader’s introduction to
the concept of knowledge, its multiple forms and uses, that it enables readers to see
the world through a gnoseological lens and empowers them to act properly, both as
individuals and as citizens.
My ideas have been influenced by numerous texts as well as exchanges of views
with colleagues and good friends. I am grateful to Cristian Calude for reading the
manuscript and for his insightful and constructive remarks.
I would like to acknowledge the role of Giorgos Hatziiakovou, Director of Armos
Publications, who believed in the book from the very outset and published its first
version in Greek. I would also like to heartily thank Daniel Webber who translated a
revised version of the Greek text into English with a lot of competence, diligence,
and talent.
Finally, I could not have written this book without the constant, undivided
support of my wife, Olga. Her love of all that is elevated and wonderful, spiritual
communion and shared experiences were catalytic in inspiring me to clarify and
formulate core ideas in the book.

Meylan, France Joseph Sifakis

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Part I For a Gnoseological View of the World

2 Fundamental Questions About Knowledge . . . . . . . . . . . . . . . . . . . . 7
2.1 Three Key Questions—Why? What? How? . . . . . . . . . . . . . . . . . 7
2.2 The Teleological “Why?” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 The Ontological “What?” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 The Gnoseological “How?” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
3 Information and Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.1 The Birth of Informatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2 What Is Information? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.3 About Computing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.3.1 Algorithms—Conventional Computers . . . . . . . . . . . . . . . 15
3.3.2 Neural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.4 Knowledge and Its Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4.1 What Is Knowledge? . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3.4.2 Types of Knowledge and Their Validity . . . . . . . . . . . . . . 19
3.4.3 Scientific Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.4.4 Technical Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.4.5 A Methodological Clarification . . . . . . . . . . . . . . . . . . . . . 28
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
4 The Development and Application of Knowledge . . . . . . . . . . . . . . . 31
4.1 The Development of Knowledge—Principles and Limitations . . . . 31
4.1.1 Modeling—The Role of Language . . . . . . . . . . . . . . . . . . 32
4.1.2 Stratification—Abstraction Hierarchy . . . . . . . . . . . . . . . . 36
4.1.3 Modularity—The “Atomic Hypothesis” . . . . . . . . . . . . . . . 37
4.1.4 Emergent Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

xi
xii Contents

4.2 Issues Related to Knowledge Application . . . . . . . . . . . . . . . . . . . 40

4.2.1 The Limits of Scientificity . . . . . . . . . . . . . . . . . . . . . . . . 40
4.2.2 Scientism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
4.2.3 Experts—The Mystification of Expertise . . . . . . . . . . . . . . 44
4.2.4 Researchers, Research, and Innovation . . . . . . . . . . . . . . . 45
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

Part II Computing, Knowledge and Intelligence

5 Physical Phenomena and Computational Processes . . . . . . . . . . . . . . 53
5.1 Scientific Knowledge and Computing . . . . . . . . . . . . . . . . . . . . . 53
5.1.1 Computational Processes and the Experimental Method . . . 53
5.1.2 Physical Phenomena as Computational Processes . . . . . . . . 54
5.2 Comparing Physical Phenomena and Computational Processes . . . 55
5.2.1 Computation and the Concept of Continuum . . . . . . . . . . . 55
5.2.2 Conflicts and Resources in Systems . . . . . . . . . . . . . . . . . 57
5.2.3 Time and Synchronization . . . . . . . . . . . . . . . . . . . . . . . . 61
5.2.4 Why Is the Physical World Comprehensible? . . . . . . . . . . . 63
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
6 Human vs. Artificial Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
6.1 Human Intelligence Characteristics . . . . . . . . . . . . . . . . . . . . . . . 65
6.1.1 Slow and Fast Thinking . . . . . . . . . . . . . . . . . . . . . . . . . . 65
6.1.2 Common Sense Intelligence . . . . . . . . . . . . . . . . . . . . . . . 66
6.1.3 Cognitive Complexity—The Boundaries
of Understanding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
6.2 Weak Artificial Intelligence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
6.2.1 The Turing Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
6.2.2 Machine Learning and Scientific Knowledge . . . . . . . . . . . 71
6.2.3 A New Kind of Knowledge—Prediction Without
Understanding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
6.3 Beyond Weak Artificial Intelligence—Autonomy . . . . . . . . . . . . . 74
6.3.1 Autonomous Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
6.3.2 Characteristic Functions and Organization of Autonomous
Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6.3.3 Should We Trust Autonomous Systems? . . . . . . . . . . . . . . 79
6.4 Artificial Intelligence—Threats and Challenges . . . . . . . . . . . . . . . 81
6.4.1 Hypothetical Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
6.4.2 Real Risks and Challenges . . . . . . . . . . . . . . . . . . . . . . . . 82
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Contents xiii

Part III Consciousness and Society

7 Consciousness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
7.1 Our “Big Bang”—Language and Consciousness . . . . . . . . . . . . . . 89
7.1.1 Language and the Boundaries of Knowledge . . . . . . . . . . . 89
7.1.2 The Mind–Brain Relationship . . . . . . . . . . . . . . . . . . . . . . 90
7.2 The Mind as a Computing System . . . . . . . . . . . . . . . . . . . . . . . . 91
7.2.1 Value Systems and Regulatory Frameworks . . . . . . . . . . . 92
7.2.2 The Components of Consciousness . . . . . . . . . . . . . . . . . . 94
7.2.3 The Mental System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
7.3 Goal Management and Planning . . . . . . . . . . . . . . . . . . . . . . . . . 102
7.3.1 Strategies and Tactics for Achieving Goals . . . . . . . . . . . . 102
7.3.2 Safety and Liberty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
7.3.3 Free Will . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
7.4 The Conscious and the Subconscious Self . . . . . . . . . . . . . . . . . . 109
7.4.1 General Considerations on Consciousness . . . . . . . . . . . . . 109
7.4.2 Learning and Creation . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
7.4.3 Creation and Creators . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
7.5 Seeking Happiness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
8 Value Systems and Society . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
8.1 About Human Societies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
8.2 Gnoseological Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
8.2.1 Epistemic Values—Truth and Falsehood . . . . . . . . . . . . . . 118
8.2.2 Technical Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
8.3 Ethical Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
8.3.1 Good and Evil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124
8.3.2 Faith and Dogmatism . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
8.4 Economic Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
8.5 The Role of Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
8.6 Depreciation of Values and Decline . . . . . . . . . . . . . . . . . . . . . . . 131
8.6.1 Decay of Language—Degradation of Concepts . . . . . . . . . 132
8.6.2 The Cycle of Lying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
8.6.3 Self-Convenience and Lack of Self-Esteem . . . . . . . . . . . . 135
8.6.4 The Triumph of the Idiots . . . . . . . . . . . . . . . . . . . . . . . . 137
8.7 About Democracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
8.7.1 Meritocracy and the Role of Leaders . . . . . . . . . . . . . . . . . 140
8.7.2 Corruption and Bureaucracy . . . . . . . . . . . . . . . . . . . . . . . 144
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
9 Epilogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
Chapter 1
Introduction

The book focuses on knowledge as useful and valid information allowing us either to
understand the world or to change the world in order to satisfy material and spiritual
needs. It provides a thorough treatment of knowledge in all its aspects and applica-
tions by machines and humans, free of ideological and philosophical bias and relying
exclusively on concepts and principles from the theory of computing and logic. As
such, it strikes a balance between popular books skirting fundamental issues and
often focusing on the sensational, and scientific/technical or philosophical books that
are not accessible to laymen.
The book progressively introduces the concept of information as an intangible
entity developed and used by both computers and living organisms. It takes a unified
approach considering the many facets of knowledge depending on their degree of
validity and truthfulness; these include implicit empirical, scientific and technical,
mathematical knowledge as well as meta-knowledge used to combine the different
types of knowledge in order to solve problems. The book also considers the different
mechanisms and processes for developing knowledge and their interrelationships by
examining their main characteristics and limitations. This provides a basis for a
discussion about the concept of intelligence and a comparison between human and
artificial intelligence (AI).
The book attempts an analysis of technical difficulties for reaching human-level
intelligence from weak AI that currently provides only intelligent systems able to
solve specific problems. It argues that the strong AI vision is not reachable by
incremental improvement of machine learning technology. The development of
trustworthy autonomous systems, as anticipated by the Industrial Internet of Things,
is considered a bold step toward closing the gap between human and machine
intelligence.
The book discusses a characterization of autonomous behavior in terms of five
basic functions and shows how this model can be adequately extended to capture key
features of human consciousness. This leads to a computational scheme that could
explain conscious behavior as a game where decision-making seeks tradeoffs

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 1
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_1
2 1 Introduction

between satisfaction of needs and the corresponding costs estimated using an

individual value system.
Finally, the book extends the model of individual consciousness into a model of
collective consciousness shaped by the action of institutions fostering a common
value system. It corroborates the thesis that human societies are information systems,
illustrated by the analysis of the underlying organizational principles as well as of
crises as manifestations of failure in their application.
The book differs from others on the philosophy of science that focus on episte-
mology (from the Greek “episteme”—science) examining the human-driven process
for acquisition of knowledge about the world. It takes a broader perspective consid-
ering all kinds of knowledge. This is why, instead, I use the term gnoseology (from
the Greek “gnosis”—knowledge), which is more appropriate, since it suggests that
knowledge is not limited to the understanding of phenomena by humans, as
discussed in the book.
The approach taken clearly separates physical reality from its models. It follows
the dualist tradition of logicians emphasizing the importance of logic and language.
In particular, the concept of information being considered is purely logical and
different from “information” in quantum physics or in the theory of communication.
From this perspective, the book mounts some hopefully well-founded criticism
toward approaches considering that information and knowledge are just emergent
properties of physical phenomena.
I have tried to explain my ideas as simply and clearly as possible while avoiding
dangerous oversimplifications. Where necessary, the text veers toward somewhat
technical, but without the knowledge required to exceed the 12th-grade level. These
small technical asides can be skipped during a first reading.
Sometimes, my love for ancient Greek literature comes through the exposition,
particularly with references and citations. I hope that this would be of interest to
international readers and would allow them insight into the sinuous course of the
human spirit in the quest for knowledge.
The scope of the issues I am discussing is such that I do not think it expedient to
provide bibliographic sources. I make certain references to websites, in particular
Wikipedia, which provide additional information and, above all, relevant biblio-
graphic sources.
The book is composed of three parts.
The first part, consisting of three chapters, presents knowledge as truthful and
useful information, as well as the processes of its development and management by
computers and humans.
The second chapter identifies three fundamental questions about knowledge:
teleological, ontological, and gnoseological. We explain that only gnoseological
questions can be approached strictly from a methodological and logical point
of view.
In the third chapter, we outline the basic concepts of informatics. We explain, in
the simplest possible terms, how knowledge is defined as information, and we
discuss key problems posed by its development and application. In particular, we
endeavor to analyze the links between scientific and technical knowledge.
1 Introduction 3

In the fourth chapter, we discuss principles of knowledge development that are

based on simplifications of reality, as well as restrictions arising from these
principles.
The second part, consisting of two chapters, analyzes the relationship between
computational processes and physical phenomena, as well as the production and
application of knowledge by machines and humans.
The fifth chapter investigates relationships between informatics and physics and,
in particular, discusses how each of these fields of knowledge can enrich the other,
comparing basic concepts and models.
The sixth chapter attempts to compare artificial and natural intelligence, trying to
answer the question of whether and to what extent computers, as they stand today,
can approach human mental functions. The chapter presents autonomous systems
that are called upon to replace humans in complex operations by realizing the vision
of strong AI. Finally, it discusses the risks—real or hypothetical—of the reckless use
of computers and artificial intelligence.
The third part, consisting of two chapters, seeks to analyze the functioning of
individual consciousness and to discuss its impact on societal organization, in the
light of the previous gnoseological perspective.
The seventh chapter explains the functioning of consciousness as an autonomous
system that manages short- and long-term goals based on value criteria and accu-
mulated knowledge.
The eighth chapter discusses how values are shaped in society and how the
composition of each individual’s subjective experience takes on an objective dimen-
sion, which can be studied as a social phenomenon. It analyzes the role of institutions
in shaping and maintaining a common scale of values for strengthening social
cohesion, as well as a discussion on the principles of democracy.
The book concludes with an epilogue summarizing key conclusions and ques-
tions on what the future holds for us.
 Part I
For a Gnoseological View of the World

We present knowledge as truthful and useful information, as well as the processes of

its development and management by computers and humans.
• We identify three fundamental questions about knowledge: teleological, ontolog-
ical and gnoseological. In it, we explain that only gnoseological questions can be
approached strictly from a methodological and logical point of view.
• We outline the basic concepts of informatics and explain how knowledge is
defined as information.
• We discuss principles of knowledge development that are based on simplifica-
tions of reality, as well as restrictions arising from these principles.
Chapter 2
Fundamental Questions About Knowledge

2.1 Three Key Questions—Why? What? How?

Our efforts to understand the world raise three kinds of fundamental questions about
knowledge.
First, teleological questions, seeking a purpose in phenomena such as “why did
the world form?” and “why do we exist?”. The use of adverb “why” implies will and
intent, which are traits of human consciousness. Therefore, the answers to such
questions cannot be controlled through reason and are beyond any experimental
validation and analysis.
Second, ontological questions, which concern the nature of “being.” For exam-
ple, “does the world exist outside our senses and our minds?” or “what precisely is
the world as such and as a synthesis of its elements?”.
These are questions about what we can know. We ask such questions because we
do not distinguish between reality and its models. These are conceptual abstractions
our minds use to understand reality. Nowadays, with quantum mechanics and
computers, it is commonly accepted that the world and its models are two separate
things. I will explain below why ontological questions are not amenable to rational
control either.
Third, gnoseological questions about how we can know, for example, how the
world is changing, how we think, how we build a building, and how birds fly.
Answers to these questions allow us to understand or to change the world. Knowl-
edge can be scientific, technical, mathematical, or purely empirical, as I will explain
in what follows. It consists of relationships that we can verify logically or validate
empirically using the appropriate methodology.
This simple classification clearly and strictly establishes a typology of fundamen-
tal questions about knowledge. As regards teleological and ontological questions, we
do not have any logical or empirical criteria to decide whether or not an answer is
sound. It is only worth attempting to answer gnoseological questions, insofar as they
are formulated as relationships between well-defined concepts.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 7
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_2
8 2 Fundamental Questions About Knowledge

It should be noted that these three types of questions are independent from each
other. Whatever your convictions on “why?” and “what?”, they do not logically
affect your knowledge about the world. Knowledge does not depend on whether you
are a Buddhist or an atheist, whether or not you believe numbers exist regardless of
the human mind.
However, to avoid misunderstanding, I must stress that I do not disregard the
importance of teleological and ontological problems in the philosophy of the indi-
vidual. It is simply that the answers each person adopts are derived from their
convictions and cannot be judged according to objective criteria.
Disputes between religion and science arise from time to time when limits are
violated, i.e., when religion asserts opinions on gnoseological issues, for example,
disputing heliocentrism or when science encroaches into metaphysical territory
without taking the necessary precautions. Certain scientists overlook inherent limi-
tations of scientific theories and the relativity of scientific knowledge or use scientific
methodology in an abusive and inappropriate manner, as I will discuss in what
follows.

2.2 The Teleological “Why?”

The most difficult questions are the “why?”s that torment us from a very young age.
“Why does the world exist?”, “why are we here?”, “why do we come and go?”, “why
are we conscious?”, and a host of others you can imagine. All of these questions are
linked to seeking purposes. We are searching for interrelationships in everything we
see and analyze. And even if we do answer a “why?”, for example, “why do
we die?”, other, deeper “why?”s emerge. This creates chains, which result in
teleological questions that cannot be answered rationally.
Plato accepts that nature was designed by an intelligent creator according to
pre-existing immutable models. The idea of teleology starts with Aristotle, on
whom Christian theologians/philosophers later relied. Aristotle criticizes Democri-
tus for his overemphasizing the necessity of physical laws but without identifying
any “final purpose”—“it is absurd to say that there is no purpose in nature simply
because we do not see it.”
It is interesting that, through historical materialism, Marx and Engels adopt a sort
of “teleological” outlook on sociohistorical processes, the difference being that they
replace the mind and consciousness with economic determinism.
The ancient Greeks took the world as it is and their questions were more of
“what?” and “how” varieties. Subsequent teleological philosophical ideas were
influenced by Judaism and Christianity, dominated by the idea that God has a plan
for humanity. Perhaps it is no accident that the term “project” does not have a precise
equivalent in the Greek language, one so rich in concepts.
In a strictly logical approach, teleological ideas boil down to rhetoric, to a
mythology that does not affect the way knowledge is sought. Of course, nothing
prevents us from teleologically interpreting the laws governing phenomena, but this
2.3 The Ontological “What?” 9

is logically superfluous. Thus, some people may ascribe a higher purpose to the law
of conservation of mass/energy, but this cannot have any rational basis.

2.3 The Ontological “What?”

Parmenides may have been the first thinker to raise the question regarding “being”
and its importance. The Parthenon is a “being,” it exists. However, I perceive it in
one way, a foreign visitor in another, an architect in another, and an archeologist in
yet another. Yet, where does its “being” lie? The fact that it is made of marble, in its
overall form and architecture, in what it depicts?
Parmenides would reject the sensible world: “he made reason the standard, and
pronounced sensations to be inexact.” He would accept as what “exists” only what is
recognized and determined by “reason” and identified with it.
To put it another way: understanding reality is reduced to a process of interpre-
tation carried out by the mind. This point of view relativizes knowledge and sets the
“substance” of reality beyond its scope. At the same time, though, it resolves certain
false problems encountered by those who have attempted such an endeavor.
The fact that each mind can see reality differently does not negate the existence of
commonly accepted views on reality. This is achieved through communication, the
formation of shared knowledge through social interaction.
The quest for an absolute substance of things and pure categories of “being” has
troubled and continues to trouble scientific and philosophical thinking. Aristotle
accused Heraclitus of being irrational for blatantly ignoring the basic logical princi-
ple of “contradiction” and the “excluded middle.”
To take an example from the more recent past, let us examine the dispute around
the particle-or-wave nature of the electron that shook up twentieth-century physics.
Many major thinkers saw a clear logical contradiction in the fact that the electron “is
simultaneously” both a particle (discrete) and a wave (continuous). Heisenberg, like
many others, would reach the hasty conclusion that “the mathematical scheme of
quantum theory can be interpreted as an extension or modification of classical logic,”
meaning a logic where the principle of contradiction does not hold true.
However, a Parmenidean view of reality rids us of these impasses and para-
doxes—as long as we give up any ambition of understanding what an electron “truly
is,” of course. Contradictions are resolved if we understand that, depending on how
we observe them, phenomena can be consistent with models that have different
characteristics, without this necessarily implying a logical contradiction. Reality is
“what it is” and we can only mentally understand reality through theories that
abstract it. The multiplicity of theories does not necessarily entail logical contradic-
tions, provided each theory takes different aspects of reality into account. Of course,
when two or more theories are applicable, an interesting problem that arises concerns
the study of the relationships between them—which was successfully done by
quantum physics.
10 2 Fundamental Questions About Knowledge

I believe that, in this way, one can understand Heraclitus’s thinking, which
proposes a dialectical perception of reality. The harmonious synthesis of opposites
does not lead to a logical contradiction. It is a paradigm for understanding reality not
as something static, but as an “evolving process.” Evolution, motion, and everything
that changes is the result of the synthesis of counteracting forces that shift their point
of equilibrium according to their balance.

2.4 The Gnoseological “How?”

The type of questions open to rational investigation is those asking “how?”. How
does a clam reproduce, how is force related to acceleration, how do you build a
house, how do you solve a mathematical problem, how do we avoid inflation? The
answers to these questions are relationships between observations involving con-
cepts and objects. They comprise what I call knowledge, which I will define more
precisely in what follows.
At this point, I should stress that these questions concern both understanding
situations and solving problems. Moreover, they concern not just the physical world,
but also abstract mental processes.
Knowledge is about questions open to answers that are general and independent
of each individual’s personal experience. It must be possible to affirm their truth in
an empirical or logical way. That is why some people believe that one can only know
the physical world and what they call “objective reality.” Such a view implies that
mental phenomena and consciousness can only be understood to the extent that they
can be studied based on the laws of physics. This often leads to the position that the
world is divided, on the one hand, into physical phenomena that are a “game” of
particles in fields of forces and, on the other hand, into the world of perceptions, of
consciousness, which is the subjective experience of each individual.
Here I agree with the philosopher John Rogers Searle, who points out that such a
distinction leads to a false dichotomy. Ontologically speaking, consciousness is
certainly a subjective experience. Death or pain are ontologically subjective experi-
ences, but gnoseologically objective, insofar as they can be ascertained in an
unequivocal way. Thus, it makes sense to study the causes of a person’s death or
the causes and treatments of pain.
A key tenet of this book is that knowledge is not just about physical reality but
also about all mental and social phenomena. As I will explain in what follows, this
implies the dual nature of reality, which consists of physical phenomena and
informational phenomena.
Chapter 3
Information and Knowledge

3.1 The Birth of Informatics

Informatics is a new field of knowledge. It was established in 1936 by British

mathematician Alan Turing, who defined a mathematical model of computation—
the machine bearing his name.
The rest is, more or less, history. Thanks to rapid advances in electronics and
materials technology, we developed increasingly powerful computers. The first
computer was built for military applications. This was followed by the use of
computers for commercial applications around the 1960s. We then connected
computers to telecommunications networks, and this paved the way for the Internet
and the Web—the information society. Another important step was the use of
computers in embedded systems to control production processes and develop ser-
vices. Today, over 95% of integrated circuits are embedded. Concealed in devices of
all kinds, they provide automated services.
At first, informatics was considered to be a technology lying somewhere between
mathematics and electronics. Over the years, it developed as an independent field of
knowledge and resisted absorption by other fields. Today, everyone agrees that it is
an important field of knowledge that extends far beyond “computer science.”
“Computer Science is no more about computers than astronomy is about telescopes,”
as Edsger W. Dijkstra used to say [1].
What we have been living through the last 60 years, the information revolution
and, by extension, the knowledge revolution, is one of the most important milestones
in the history of humanity.
Since antiquity, people have been building machines to multiply their muscular
power, to set up the technical civilization that has so radically changed the world.
Thanks to computers, people are able to staggeringly multiply their mental abilities.
This is because computers far surpass humans in terms of speed and accuracy of
computation, while creativity and the ability to understand the world remain the
privilege of humans. There is, thus, considerable complementarity.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 11
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_3
12 3 Information and Knowledge

We are at a great landmark in humanity’s path toward conquering knowledge,

which began with the appearance of language, the discovery of writing, and the
invention of printing.
Despite its importance, informatics has not been accordingly acknowledged in
primary and secondary education. It does not receive the same emphasis as physics
or biology. It is taught more as a form of technology, without students delving into
theories of computation. Public opinion and decision-makers do not even grasp its
importance as a field of knowledge that can fundamentally change not only our
perception of the world but also the very existence of human beings.
The truth is that the so-called exact sciences dominated scientific thinking until
the end of the twentieth century. As an academic, I had the opportunity to converse
with famous scientists and was often surprised to realize a lack of understanding of
the importance of informatics. It is true that scientists try to explain the world and
everything complex as a combinatorial game of particles, the whole as a synthesis of
its parts. This reductionist approach wants complex phenomena to “emerge” from
the synthesis of simpler ones.
As Richard Dawkins characteristically notes [2]: “My task is to explain elephants,
and the world of complex things, in terms of the simple things that physicists either
understand, or are working on.” However, this approach fails to explain what
information is and what computation is. At times, it leads to ridiculous assertions,
such as Stephen Hawking’s claim that “the brain could exist outside the body” [3]. In
other words, he believes that he could go on living after death by placing his brain’s
information in a memory stick.
Such statements suggest a deep ignorance of what information is, of how
cognition and consciousness work. It is a pity that they strike a significant chord
with the media.
I would like to explain as simply as possible what the nature of information and
computation is. Needless to say that at present, there is still no full agreement on
these concepts, even among computer scientists.
The debate on the relationship between informatics and other fields of knowledge
raises deep and delicate epistemological problems. My aim is not to provide answers
to these problems, which will remain unresolved for a long time. I simply propose a
methodological framework for comparing informatics to other fields of knowledge. I
will also explain what can be achieved with computers today, what their inherent
limitations are, and how much they can contribute toward developing fruitful
interdisciplinary cooperation.

3.2 What Is Information?

Informatics involves the study of computation as a process for transforming infor-

mation—what we can compute and how to go about computing it.
Yet, what is information?
3.2 What Is Information? 13

Fig. 3.1 Information as a relationship between symbols and concepts

Let me first give an example. When I see the symbols 4, 100, δ, IV, I can interpret
them as the concept of “four” in the decimal system, in the binary system, or the
ancient Greek and Roman numeral system, respectively. What the symbols represent
is a matter of convention. The interpretation of the information codified in these
symbols depends on whether the mind of the person seeing the symbols is familiar
with this convention. Familiarization can take place either through empirical learn-
ing or through education. For all non-color-blind people, the color red corresponds to
the concept “red,” and the photograph of an apple recalls the concept of an “apple.”
Conversely, Maxwell’s equations constitute information only for those familiar with
electromagnetism. The Linear A script, insofar as it has not been deciphered, does
not convey information.
Information can be defined as a semantic relationship between a symbolic
language and a set of concepts. A symbolic language is described by a set of basic
symbols, usually an alphabet and rules that specify how symbols are combined
(structured) to form more complex elements, i.e., words and phrases. One common
example is natural languages. The same concept is represented differently in differ-
ent languages. The information of the various representations is precisely the
relationship with that concept. Of course, in addition to natural languages, there
are also programming languages, mathematical languages, the language of images
expressed in pixels, sign language, etc.
How is information created? Firstly, one must define the symbolic language, its
symbols and its syntax, that is, the set of rules governing the combinations of
symbols required to form complex entities, such as words and phrases. The defini-
tion does not need to be rigorous and can emerge experientially as a practical result.
Nevertheless, it requires two relationships, as shown in Fig. 3.1.
One codifies concepts and provides their representation through symbols.
The other provides the meaning or semantics of the language. It decodes a
symbolic representation by associating it with concepts. It should be noted that
this relationship must be compositional: the meaning of a phrase is determined by the
meanings of its constituent words and its structure. Therefore, the meaning of the
phrase “I give the book to George” arises by composing the meanings of “I,” “give,”
“the book,” and “to George,” taking into account the grammatical analysis of the
structure that identifies the subject, the verb, and the two objects.
14 3 Information and Knowledge

For the time being, I will not try to explain what the mind is and exactly what
concepts are. It is something I consider as given—just as energy and time are
considered as given, albeit recondite, in physics.
Now I would like to stress certain characteristic properties of information, and
explain that, as an entity, it is different from physical entities.
• Unlike the basic entity of the physical world, which is matter/energy, information
is an abstract relationship defined independently of the space-time. All our
concepts and knowledge of the world are space- and time-independent, for
instance, a mathematical function computed by a program. Of course, the execu-
tion time of the program depends on how fast the computer is—but that is
irrelevant to the nature of the computation itself.
• One characteristic attribute of information that it is intangible, i.e., it is not subject
to the laws of physics. While it needs a medium (matter or energy) to be
represented, such as sound, image, or electrical waves, it is independent of the
medium used and its properties. For example, when I send an e-mail, it can be
read by the recipient and transmitted orally to a third party, who can place it
within a text, etc. The information is the content of the message. What matters in
this process of transmitting information is for the content to remain unaltered,
whatever the encoding and mode of transmission.
Years ago, when I gave a lecture to gifted children in India, a young boy raised his
hand and asked me: “Sir, does information weigh anything?”. I asked him, “What do
you think, if you erase your computer’s memory, will its weight change?”. He
correctly answered “no.” This also reminds me of a joke where a father and his
son are eating Swiss cheese, the one full of large holes. The boy asks, curiously,
“Dad, the cheese we eat goes to our stomach, but can you tell me what happens to the
holes?”. That is precisely what information is: it is the structure, the organization of
a medium to which we can ascribe meaning.
Computational phenomena are transformations of information that may result in
the creation, loss, or preservation of semantic content—which is a more technical
matter, so let’s discuss it later. A key axiom in my view is that creating information is
the exclusive prerogative of the mind. Computers cannot generate information.
Algorithms are transformations that preserve information defined by programmers.
While information is intangible, it matters far more than tangible goods in modern
economies. That is why we speak of an intangible or a knowledge-based economy.
It is the information content of our brain that determines to a large extent who we
are. Imagine that, for a moment, your memory was erased. Externally, you would be
the same person, with the same physical attributes, but essentially you would no
longer be “you.”
In conclusion, I must point out that the concept of information as I present it is the
one defined by Alan Turing, and it is foundational for informatics. It must not be
confused with another concept of quantitative information, what we usually call
“syntactic information,” which concerns the minimum quantity of memory or any
medium necessary to represent or transmit information.
3.3 About Computing 15

I refer to Shannon’s theory [4] and other similar theories that determine a measure
of the information content of a structure, irrespective of what it means. These
theories are useful for the information economy, i.e., to assess how much the storage
(memory) or transmission (channel bandwidth) of information costs in terms of
natural resources. In contrast, it is not at all useful in helping us understand what
computation is.
Thus, if in the sequence of symbols “good morning” I change the order of the
letters, I obtain sequences with the same syntactic information, regardless of what
they mean. It is like saying that a kilogram of cotton and a kilogram of gold are
equivalent because they weigh the same!
Unfortunately, many authors explain the concept of information in this way as a
physical quantity and do not care about its correlation with languages, and comput-
ing. They thus adhere to the “brave” tradition of those who turn a blind eye and,
when unable to describe something according to their favorite theory, prefer to
ignore it rather than to question the theory’s suitability.

3.3 About Computing

3.3.1 Algorithms—Conventional Computers

An algorithm is a process or set of rules to solve a problem, such as baking a cake,

estimating the temperature of the planet or buying a book online. It is convenient to
think of an algorithm as a procedure that allows for the calculation of mathematical
functions. An algorithm that calculates a function f is a sort of “recipe” that says, for
a given value of x, how one can calculate the corresponding value f(x). Of course,
computers run algorithms “blindly” and apply symbol transformations without
“understanding” their meaning.
For technical reasons, the representation of information in computers is binary,
i.e., the words of the computer language are written as sequences of two symbols
commonly represented as “0” and “1”.
Figure 3.2 shows the execution of an algorithm that calculates the “sum of two
integers” function for the values 5 and 7. The algorithm starts with the binary
representation of 5 and 7. It then calculates the result in successive steps, which
once decoded, gives us the number 12. The arrows show how the algorithm
combines the respective powers of two of the binary representations of the numbers
5 and 7 to provide us with the binary representation of the number 12. We usually
write algorithms in programming languages. A program that describes an algorithm
is a sequence of instructions that we can run on a computer.
I will not elaborate on my explanations concerning algorithms, but I would like to
comment on two fundamental principles of the theory of computation.
The first principle follows from theorems proven in 1931 by Austrian mathema-
tician Kurt Gödel, which show that, among all mathematical functions, there are
16 3 Information and Knowledge

101 + 111

1+1+1 0+1+1 1+1

11 10 10

1100
Fig. 3.2 The concept of algorithm as a symbol transformation process

functions that are not algorithmically computable. In fact, the set of non-computable
functions is non-enumerable, i.e., far larger than that of computable functions, which
is enumerable. This result sent shockwaves through the world of mathematics, as it
clearly limited our ability to accurately calculate a huge number of functions and to
solve mathematical problems. Unfortunately, many non-computable functions are
very useful, such as those that allow us to check whether the programs we write are
correct. For example, the veracity of the phrase “a program will terminate” or “the
computer memory is sufficient to run a program” is not computable.
Gödel’s results have important practical implications that will be discussed in
what follows. Their immediate consequence is that we cannot use computers to
check the correctness of the software we write. This theoretical limitation constitutes
a type of “uncertainty principle” for informatics. Some non-computable functions
can be approximated and, of course, as the accuracy of computation increases, so
does the cost of the calculation.
The second principle stipulates that there is a cost for running an algorithm
characterized by its computational complexity. The latter is the quantity of resources
(memory and time) needed to execute the algorithm. Of course, the same problem
may be solved using algorithms with a different complexity, but their complexities
cannot fall below a certain threshold. In other words, just as there is friction in nature,
meaning that when we produce work, we inevitably have energy losses, we must
also consume resources when running programs. The worst case of complexity is the
one where the cost of computation in terms of time or memory is an exponential
3.3 About Computing 17

function of the size of the initial data. However, even algorithms with relatively low
complexity can be practically incomputable.

3.3.2 Neural Networks

Here I must point out that there are other computational models that are completely
different from the conventional ones used by our computers. These models mimic
natural computation processes, and there is no clear separation between the algo-
rithm/program and the computer running it. Such models include those of natural
computing, which we will discuss in what follows and which leads to different kinds
of computers, such as analog, quantum, and neural networks. The latter adhere to a
computational model that mimics the neural networks of our brain.
While the principles of neural networks have been known to us since the
mid-1940s, advances in research have enabled them to be applied with marked
success to problems of artificial learning over the last two decades, giving a huge
boost to the development of artificial intelligence and its applications.
Neural networks allow for the effective solution of problems whose algorithmic
solution is hampered by an insurmountable complexity barrier. Let us take a simple
example by assuming that I want to build a system where I input images of cats or
dogs, and the system correctly identifies the animal being depicted.
The traditional algorithmic approach starts with an analysis of models that
characterize each animal’s form with a set of patterns, such as the shape of the
animal’s head, the position, and shape of its eyes, ears, and nose. Based on this
analysis, I would have to write an algorithm that analyzes images, identifying
characteristic patterns, and deciding whether the image is a dog or a cat. This
model-based algorithmic approach is very difficult to implement when analyzing
composite images and is computationally very costly.
Machine learning takes a completely different, empirical data-based approach
that does not require model analysis and programming. A neural network is a system
that, for each information being input, generates a corresponding response. The
network learns to distinguish cats from dogs through an experimental process, just
like a child. In other words, we “show” the system a large number of related images
and progressively configure its parameters so that it provides the correct response.
The “training” process of the system is adaptive: the percentage of incorrect
responses decreases as the size of the training dataset increases.
Of course, a child learns with just a few examples, while training neural systems
requires a large volume of data. However, their advantage over traditional systems
running algorithms (programs) is that, after potentially time-consuming “training,”
neural systems compute their response almost instantly. As a result, companies such
as Nvidia and Waymo have developed neural systems that are capable, after
intensive “training,” of driving cars with relatively small but still non-negligible
error rates.
18 3 Information and Knowledge

3.4 Knowledge and Its Validity

3.4.1 What Is Knowledge?

Knowledge is information which, incorporated into the right network of semantic

relationships, can be used either to understand a situation or to act in order to achieve
a goal.
This definition considers knowledge to be useful information, and it should
therefore have a certain degree of truthfulness and validity. The phrase “the sun
revolves around the earth” is information, but it is not knowledge, even though it was
considered to be knowledge in the past. A recipe for baking a pie or the solution to a
mathematical problem is knowledge insofar as it enables us to achieve
corresponding goals.
It is important to understand that knowledge, as we have defined it, has a general
and dual character (Fig. 3.3).

Fig. 3.3 Knowledge as information and its dual use

3.4 Knowledge and Its Validity 19

• On the one hand, it allows us to understand situations, i.e., what is happening

around us, which includes the physical world, societies, and their artifacts. It can
be a simple item of information that holds true at a given moment in time, such as
a measurement (my maximum blood pressure is 14), the statistical result of a
survey poll (45% positive opinion) or experimental knowledge (the speed of
sound is 343 m/s). This type of knowledge includes both scientific and shared
empirical knowledge.
• On the other hand, it allows us to achieve a goal. In this case, knowledge is linked
to meeting needs and entails action to change the state of the agent or of its
external environment. This type of knowledge includes technical knowledge,
such as that of an engineer (in order to design a house) or a programmer
(to write a program). It also includes implicit empirical knowledge that allows
us, for example, to walk or to speak.
Below, I propose a classification of the different types of knowledge according to
their validity and applicability. I also explain that there are different ways of
developing knowledge that also determine how it is transmitted and used.
At this point, I would like to emphasize a fundamental tenet of this book. The
world can only be understood as the combination of phenomena of the material
world and its mental representations. This dualist vision of the world distinguishes
between two basic entities: matter/energy and information, corresponding respec-
tively to what can be observed in the material world (phenomena) and what can be
thought (noumena). This is a very important distinction so as not to confuse, as many
often do, “the map with the country,” i.e., the world with its models. In this context,
the human mind plays a key role. It is a “supercomputer” that either instantly creates
knowledge by explaining the world or takes action by applying knowledge in order
to achieve goals.
The above position differs from others, which consider matter/energy as the only
fundamental entity and, therefore, mental phenomena as emergent properties (epi-
phenomena). I hope to demonstrate in the following chapters that such positions
have inherent weaknesses. They are unable, taking a bottom-up approach, starting
from the particle world, to explain mental phenomena, such as language and
consciousness, as emergent properties of the brain.
Similar problems also arise with positions that prioritize an intangible entity such
as information, taking a top-down approach.

3.4.2 Types of Knowledge and Their Validity

The issue of the validity of knowledge has been much debated and is very timely
because knowledge is produced and used not only by people but also by computers.
Many believe that there is a wall between scientific and non-scientific knowledge.
However, I will show that subtler distinctions may exist.
20 3 Information and Knowledge

Fig. 3.4 The knowledge pyramid—classification of types of knowledge

Figure 3.4 proposes a hierarchical classification of the different types of knowl-

edge produced by humans or machines.
One significant difference is that between empirical and non-empirical
knowledge.
Empirical knowledge is acquired and developed through our senses based on
observation and experimentation. It includes all the knowledge we have gained
automatically through our experiences or after informed mental processes.
Non-empirical knowledge, which is also called a priori knowledge, is the product
of mental processes that are not directly related to our experiences. Its validity
depends only on the consistency of our reasoning. It expresses relationships that
remain unaltered and independent of space-time, where physical phenomena occur.
It includes, among other things, mathematics and logic, as well as theories of
computation. Concepts such as straight line, plane, and unit, constitute a priori
knowledge. The Pythagorean theorem or Gödel’s theorems are “eternal” truths,
propositions whose validity depends only on the axioms of Euclidean geometry
and arithmetic, respectively.
A higher kind of empirical knowledge is that based on mathematical models. This
includes scientific knowledge for a reasoned understanding of the world and techni-
cal knowledge for the safe and effective achievement of goals. These types of
knowledge are discussed in detail below. They differ from common empirical
knowledge in that they constitute a generalization of observations embodied as
mathematical models. This allows for an explanation of the relationships (laws)
between variables representing the observed quantities, for example, that
3.4 Knowledge and Its Validity 21

acceleration is proportional to force or that the drag force of a moving object is

proportional to the square of its speed. The existence of such relationships, which we
call physical laws, also allows for what we call predictability. Once there is a law
characterizing a phenomenon, it is possible to predict the impact of the change of
some variables (causes) on its overall dynamics.
The simplest type of empirical knowledge is facts and situations in a given place
and time. For example, the phrase “the temperature in Athens today is 25 degrees”
describes a situation, while the phrase “the Battle of Waterloo took place on Sunday,
18 June 1815” is a fact. This type of knowledge is of limited applicability but is
necessary to understand the world.
Most of our knowledge is common empirical knowledge. Much of our intelli-
gence relies on it. It is the result of a generalization of empirical knowledge, which
may take place consciously or subconsciously. Its validity is simply based on its
practical use.
It is thanks to this kind of knowledge that we can walk, speak, play instruments,
dance, etc. It also includes all skills and knowledge based on common sense, such as
implicit relationships we consider “obvious.” We therefore know that parents are
older than their children or understand the dynamics of the kinetic states of objects,
even if we do not know the laws of physics (see Sect. 6.1.2).
I include the data-based knowledge generated by computers, such as that pro-
duced by neural networks, in common empirical knowledge. It is clearly empirical
and of the same nature as the common empirical knowledge of humans. It is
knowledge that allows for predictability, as a neural net computes a relationship
between inputs (causes) and outputs (effects). However, it does not have the validity
of scientific knowledge, insofar as we cannot generally describe this relationship
using mathematical models, as I explain below.
Finally, knowledge generated through data analytics techniques can be either
empirical or scientific when they are based on statistics theory.
At the top of the pyramid is meta-knowledge, which some call “wisdom.” It is
knowledge on managing any form of acquired knowledge. It allows us to combine
knowledge of all kinds. It includes problem-solving, design, and decision-making
methods. It also includes non-formalized knowledge used in professional skills.
The above classification demonstrates that, depending on how it is developed,
knowledge has different degrees of validity and generality. It allows us to understand
the differences between natural and artificial intelligence, which will be discussed
extensively in the following chapters.

3.4.3 Scientific Knowledge

3.4.3.1 The Process of Developing Scientific Knowledge

Scientific knowledge helps us understand the physical world. It is mainly based on

analytical thinking, which connects phenomena to the world of mental models,
22 3 Information and Knowledge

information, and mathematics. As I have said, its development leads to the formu-
lation of laws, unalterable relationships that govern the changes observed, for
example, conservation of energy, universal gravitation, gas laws, and the law of
supply and demand. It follows three steps whose realization raises different kinds of
problems.
1. The first step aims at capturing the observed reality through a model. For
example, if I want to study meteorological phenomena, I will first build a
model based on a number of empirical assumptions. The model may be formal-
ized in a purely mathematical manner, for example, a system of equations, or in a
less rigorous and ad hoc manner. It expresses relationships between quantities
such as pressure, temperature, direction, and speed of winds. Of course, I can
create different—and not necessarily comparable—models for the same phenom-
enon. This is determined by what is regarded as observable and measurable.
The difficulty of modeling a phenomenon is characterized by what is called its
epistemic complexity, i.e., how difficult it is to find an appropriate mathematical
or mental model for the phenomenon we are studying. For example, let us assume
that someone is experimenting in an effort to understand a phenomenon where a
quantity increases exponentially. If the experimenter was, for instance, the
ancient mathematician Thales of Miletus, who only understood proportionality
relationships, then he would not be able to explain the phenomenon.
We know that, in order for Newton to formulate his theory, he had to develop
the relevant mathematical models (differential calculus). Without these models,
the discovery of the theory of universal gravitation would have been impossible.
By this, I simply mean to say that the scientific laws we discover depend on the
models at our disposal. There are extremely complex phenomena for which we
cannot find suitable models, thus limiting the possibilities of advancement of
scientific knowledge. For example, understanding human behavior is faced with a
practically insurmountable “wall” of epistemic complexity.
Note that sometimes the inability to study complex phenomena is not so much
due to a lack of theory, but to our inability to formulate a model that faithfully
represents the phenomenon by combining a large number of parameters and facts.
For example, let us consider a system with two teams of robots playing
football. I could theoretically predict the outcome of the game if I could build a
faithful model of the system: accounting for each player (as an electro-mechanical
system with hardware and software), for the objects in its environment and how
they interact. The explosive difficulty of the problem does not lie in studying new
phenomena or discovering new laws. It concerns the faithful description of all the
extremely complex relationships that characterize the dynamics of the
phenomenon.
2. The second step in the discovery of scientific knowledge concerns the analysis of
the model, carried out by experts usually assisted by computers. The analysis will
reveal relationships between the observed quantities of the model, either in the
form of explicit mathematical relationships, for example, cause-and-effect rela-
tionships, or in the form of graphical representations of the phenomenon, usually
3.4 Knowledge and Its Validity 23

generated by computers. Thus, an accurate weather prediction achieved by

resolving a meteorological model depends both on the model’s faithfulness and
on the accuracy of the resolution method.
Thanks to the use of computers, we can analyze complex models and push the
boundaries of knowledge as far as possible. At this point, I must recall the
limitations concerning computational complexity (see Sect. 3.3.1). If a precise
resolution of the model is not feasible, then we may be able to find approximate
solutions that are satisfactory in practice. The higher the accuracy of resolution of
a model, the better the predictability of the phenomenon.
I noted above that there might be more than one model for the same phenom-
enon, representing it in different ways or in different degrees of detail. Detailed
models primarily allow for better predictability, but their analysis may involve
increased computational complexity.
3. The third step in the discovery of scientific knowledge is to verify the faithfulness
of the model. This means that if a relationship holds true for the model, for
example, a prediction concerning tomorrow’s weather, this will also be con-
firmed, with a certain tolerable error margin, by observing the phenomenon
under study. Of course, confirmation requires comparison of a sufficiently large
number of experimental data and an investigation demonstrating that the data
considered “sufficiently” cover a representative set of states of the phenomenon.
This raises certain methodological problems, as physical phenomena have, in
theory, an infinite number of states that cannot be investigated. Error is therefore
logically possible, however large the number of experimental confirmations.
This context gives rise to the problem of exploring the entire spectrum of states
in the best possible way and concerns the controllability of the phenomenon
under study. It is the possibility of bringing it into predefined initial states, which
is not feasible, for example, for numerous geophysical, astrophysical, economic,
and social phenomena. The distinction between controllable and uncontrollable
phenomena creates a dividing line as to the validity of scientific truth. In classical
physics, phenomena are largely controllable. We can reproduce the phenomenon
starting with the same initial conditions and make it repeatable, which confirms
that our model’s predictions are not random.
There are many phenomena that we cannot reproduce for various reasons.
Some, such as social phenomena, are not repeatable by their very nature. Some
phenomena could be reproduced with some probability, while others are difficult
to bring to exactly the same initial conditions, for example, economic systems.
For a scientific field, reduced controllability implies decreased validity of knowl-
edge. Nonetheless, this limitation should not imply any disparagement of the
field.
In summary, applying the scientific approach and producing scientific knowledge
involve three steps and corresponding complexities (see Fig. 3.5).
1. Epistemic complexity, which depends on our ability to grasp a phenomenon
using existing mathematical models and modeling tools.
24 3 Information and Knowledge

Requirements

Fig. 3.5 Development of scientific and application of technical knowledge

2. Model resolution complexity, which is mainly computational and depends, in

turn, on the complexity of models and the effectiveness of resolution techniques.
The latter is determined by computer performance and the availability of suitable
model resolution algorithms.
3. Validation complexity, which depends on the degree of controllability and the
dynamics of the phenomenon under study.
I would like to stress that the proposed methodology is more general than others
limiting the development of scientific knowledge to cases where phenomena are
controllable and can be experimentally reproduced. I believe such a limitation makes
no sense. On the one hand, the degree of controllability of phenomena may vary. On
the other hand, it denies entire fields of knowledge, such as social sciences, scientific
status.

3.4.3.2 On the Nature of Scientific Knowledge

Having explained the process of developing scientific knowledge, I would like to

highlight some of its features.
Our knowledge of physical, economic, and social phenomena is described by
laws which express relationships between observed quantities in space-time. These
3.4 Knowledge and Its Validity 25

“laws” are not of the same nature as the laws of a legislative system, whose violation
entails criminal liability. Nature is not responsible if it does not comply with what we
consider to be its laws.
We have seen that the development of scientific knowledge depends on the ability
to find appropriate models and concepts provided by the language of mathematics
and logic. Therefore, the laws of phenomena are “invented” by people rather than
“discovered.” Some people wish to believe that these laws predate human thought.
This is an ontological issue, and I will not discuss it further.
Laws are created in the mind of the observer as a generalization of relationships.
In other words, by ascertaining that a mathematical relationship is validated between
observed quantities for a large number of experiments, we declare this relationship a
law. Such a leap in logic involves arbitrariness, as no validation can be made
regarding the infinity of possible combinations of values. For a set of values, the
observed relationship could be falsified—and this was indeed the case when the
theory of relativity demonstrated that Newton’s simple and elegant theory of uni-
versal gravitation is not valid for high velocities.
This generalization is called physical induction and is different from mathemat-
ical induction, which is based on axioms and produces mathematical knowledge (see
Sect. 4.1.1.2 Formalized Language—Theory).
Scientific knowledge therefore states, with some degree of certainty, what can
happen “here and now.” No matter how many observations satisfy the inductive
relationship, there is no guarantee that the next observation will not violate it, thus
definitively canceling its validity.
The renowned mathematician and philosopher Bertrand Russell derides the
confidence of physicists regarding the power of scientific truth with the story of
the “inductivist turkey” [5]. It is about a turkey on an American farm that decided to
construct a scientific view of the world in which it lives. Thus, on its very first day on
the farm, it observes that it is fed at 9:00 in the morning. Of course, it did not rush to
draw conclusions. It waited for the next day and found that it was at 9:00 in the
morning, and then every day of the week, rain or shine. Therefore, ultimately
satisfied and considering that it had a fairly large number of confirmed observations,
it inductively concluded that “I am fed every day at 9:00 in the morning”—that is,
until Christmas Eve, when, instead of being fed, it had its throat cut.
Unfortunately, the relativity of scientific knowledge is yet to be sufficiently
emphasized or understood, even by famous scientists. What is one to think of the
assertion of the famous physicist Stephen Hawking who believed that “the universe
and the “Big Bang” was an inevitable consequence of the laws of physics” [6]?

3.4.4 Technical Knowledge

Technical knowledge uses synthetic thinking. It is used to build artifacts that meet
given techno-economic goals to satisfy human needs.
26 3 Information and Knowledge

There are many examples of complex systems challenging our technical knowl-
edge. One, for instance, concerns autonomous transport systems and self-driving
cars, in particular, which we hope to see in a few decades. Another “distant” system
is the “Next Generation Air Transportation System” which envisions the complete
automation of air traffic control [7].
The development of technical knowledge for building artifacts consists of three
steps, the realization of which raises different kinds of problems (Fig. 3.5).
1. The first step is to collect, understand, and formalize the needs leading to the
design of the artifact. Needs are initially expressed as a set of requirements in
natural language. After having been analyzed, requirements are organized by
experts into technical specifications, which form part of the contractual agreement
between the system developer and the stakeholder (client). This text, many
hundreds of pages long in cases of complex systems, must be checked according
to two criteria: (a) logical consistency, i.e., that it does not contain contradictory
statements; and (b) completeness, i.e., that the specifications adequately cover all
the required operational needs. This check is carried out by specialized engineers
aided by computers. However, its automation is hampered by difficulties inherent
to formalizing and analyzing natural languages.
For some application areas from the realm of physics, the formalization of
needs may be relatively easy, for example, for an electrical circuit, insofar as its
technical characteristics can be formulated using existing theories. For other
applications, for example, flight control systems, the problem is harder as the
precise formulation of their properties requires mathematical languages that are
complex and difficult to use.
Another difficulty is the comprehension of needs of users involving criteria
that cannot be objectively defined. For example, functional and esthetic require-
ments that are difficult to formalize may be added to the technical specifications
of a house.
The difficulty in standardizing needs is characterized by what we usually call
conceptual complexity, i.e., the difficulty of formulating and structuring descrip-
tions in a given language. It is obvious that the formulation of the technical
requirements of a bridge is far simpler than those of a self-driving car. In the first
case, we can amply rely on concepts from mechanical engineering, while in the
second we must address, in addition to technical issues, other badly formalizable
issues such as traffic regulations and even legal aspects, for example, the driver’s
liability.
Once needs have been definitively formalized, the experts proceed with
designing the artifact by following methodologies that vary depending on the
application sector. Methodologies determine the organization of design work
carried out by engineering teams assisted by computer-aided design tools.
The aim of the overall process is to design an artifact in the form of a model
that meets the initial requirements. Such models include the blueprints of a civil
engineer, the wiring of an electrical system, or the software of a computing
3.4 Knowledge and Its Validity 27

system. Since the process cannot be automated, there is always a risk that the
result does not meet expectations, which we will discuss later.
2. The second step in the development of technical knowledge for building artifacts,
concerns using the design model to synthesize a construction model that sets out a
manufacturing process. The synthesis matches elements of the design model with
functionally equivalent material components, taking into account their physical
properties such as their strength and resilience. We can choose different qualities
of materials for the same design, provided, of course, we meet the original
technical requirements. This is why the synthesis also involves design space
exploration aiming at optimizing construction costs. Based on the construction
model, we can determine the techno-economic characteristics of the artifact, such
as construction costs, safety, and performance.
Depending on the technical area, the synthesis of the construction model can
be largely automated. Thanks to computers, we can build complex aircraft, trains,
factories, enormous bridges, and buildings. However, the most complex artifacts
at present are undoubtedly computerized systems, whose design and construction
would be impossible without the help of computers.
3. The third step in the development of technical knowledge for building artifacts,
concerns validating that the construction model meets the original requirements.
The approach is similar to that for validating scientific truth, except that it is
usually not carried out on the artifact but by simulating the construction model.
Once again, the problem of covering and investigating a large number of states
arises. Fortunately, the use of simulation techniques allows for better controlla-
bility. There are various validation techniques, the most commonly used of which
is testing, which involves an experimental analysis of the behavior of the artifact:
application of scenarios and comparison of behavior with the one envisaged in the
initial requirements.
We have described the general principles for the development of technical
knowledge in order to build artifacts. Three types of obstacles and corresponding
complexities also arise in this case (see Fig. 3.5):
1. Conceptual complexity, which expresses the inherent difficulty in describing the
technical specifications for the artifact to be built.
2. Complexity of synthesis of the construction model, which is mainly the com-
plexity of computing optimal construction solutions.
3. Complexity of validation, which is mainly the complexity of exploration and
coverage of a sufficient number of states confirming that requirements have
been met.
We will demonstrate in the next chapters that the difficulties vary depending on
the technical domain. For some sectors, there is no need for validation at all, while
for others, such as computer systems, validation costs account for a significant part
of the development cost, for example, over 50% for large systems.
I would also like to point out that the development of a relatively complex
information system involves significant risks of partial or complete failure. If we
28 3 Information and Knowledge

measure complexity in lines of code, the possibility of failure in the development of a

system with over one million lines of code, such as an operating system, is approx-
imately 30%, while the probability of partial failure is over 50%. Interested readers
can find numerous relevant sources online, such as [8].

3.4.5 A Methodological Clarification

To conclude this chapter, I would like to point out that I adhere to a definition of
knowledge that is broader than that found in philosophy books and dictionaries,
where it is usually equated with scientific knowledge. It is worth noting that in
English the field of philosophy that deals with the study of knowledge is called
“epistemology,” a term that relates to knowledge for understanding the world,
particularly scientific knowledge.
The Oxford Dictionary provides the following definition for “science” [9]: The
intellectual and practical activity encompassing the systematic study of the structure
and behaviour of the physical and natural world through observation and
experiment.
Other dictionaries provide similar definitions. Their common feature is that
science is interested in discovering facts and laws that govern the world. Thus,
while physics and biology are considered sciences under that definition, mathematics
and informatics are not pure sciences. This is because, as I explained, even though a
large portion of mathematics was developed in order to study physical phenomena,
mathematical truths are completely independent of these phenomena. I would say
that the same holds true for the principles of informatics. Furthermore, this definition
does not include anything related to the application of scientific knowledge—for
example, engineering or medicine are not pure sciences.
Note that the acceptance of a narrow definition of knowledge that links it to
science alone has occasionally led to sterile discussions.
Another problem with the current definitions of science is that they overstress the
importance of the experimental method. Thus, some people doubt that economics or
psychology are sciences. Others consider that there are impermeable limits between
science and non-science.
In order to avoid pointless disputes, I believe that knowledge has degrees of
validity, which depend on whether it is amenable to experimental confirmation and
the degree to which it fits reality. Observation of astronomical phenomena offers
limited possibility of repeatability. However, such restrictions challenge us to devise
methods that will push the boundaries of knowledge further and improve its validity.
I believe that we should get rid of narrow definitions that call into question the
uniform nature of knowledge. We must focus on the concept of knowledge in all its
forms, as it is connected to computing, that is, information that is useful for solving
problems, a concept that embraces scientific knowledge, technical knowledge,
mathematics, and general empirical knowledge, which becomes particularly impor-
tant through artificial intelligence.
3.4 Knowledge and Its Validity 29

As for those who overstress the importance of scientific knowledge and its
“superiority” over technical knowledge, I would remind them that their development
has been concurrent and complementary over the centuries. The ancient Greeks
clearly distinguished “episteme”—science from “techne”—art, and the latter had a
broader meaning than it does today, initially referred to skillfulness accompanied by
knowledge in order to perform a work or profession. Great mathematicians and
physicists such as Archimedes and Heron of Alexandria were famed engineers.
Furthermore, the advancement of mathematics and physics during the Renais-
sance was largely carried out by engineers such as Galileo and Leonardo da Vinci.
Today more than ever, scientific knowledge and technical knowledge are
intertwined in a virtuous cycle of mutual interaction and growth. Scientists build
complex experiments to study natural phenomena, while engineers need increas-
ingly sophisticated theories in order to build systems.
To sum up, I would like to stress that knowledge should not be compartmental-
ized. Knowledge has varying degrees of validity and multiple ways of development,
especially thanks to computers. Advances in medicine and biology, physics and
chemistry have been astounding. Even mathematics and logic, an exemplary form of
abstract knowledge, acquire an experimental dimension thanks to computers, which
are used to prove theorems and establish results.
Modeling and simulation techniques enable the study of complex physical phe-
nomena, such as the flow of lava or the behavior of complex information systems.
Models might not be directly related to observation and experimentation; they may
be entirely ad hoc or may combine theoretical and empirical results. What truly
matters is whether the results are consistent with the observations of the phenomena
being simulated, and whether they allow them to be understood and predicted.
Contrary to the prevailing view in physical sciences, the development of knowl-
edge does not always start with the observation of phenomena and experimentation.
This holds true for mathematical knowledge, but sometimes for scientific knowledge
as well. The theory of relativity indeed began with observations but was developed
in a theoretical framework based on thought experiments that were validated by
observation at a later stage. The theory of computation is based on a priori mathe-
matical knowledge. A doctor uses complex technology to diagnose a disease,
making assumptions that could explain the symptoms. This is followed by a trial-
and-error process in order to reach the most well-founded diagnosis. Failure to
comply with the “classic” methodology of developing knowledge in the physical
sciences does not logically imply invalidity.
As explained in the Introduction, instead of the term “epistemology,” I prefer the
term “gnoseology,” which also exists but is less prevalent. However, it is more
appropriate as it avoids the misunderstanding that knowledge is limited to
comprehending phenomena. I do, of course, employ the term “epistemology”
when referring to the study of scientific knowledge.
In conclusion, I would like to once again emphasize that knowledge today is more
important and strategic than material goods. It allows for dominance over the
physical world and response to global challenges, provided, of course, we possess
the meta-knowledge to manage it properly.
30 3 Information and Knowledge

References

1. https://en.wikipedia.org/wiki/Edsger_W._Dijkstra
2. The Blind Watchmaker, Richard Dawkins, 2015.
3. “Brain could exist outside body – Stephen Hawking”, The Guardian, Sept 23, 2013.
4. https://en.wikipedia.org/wiki/Information_theory
5. https://mashimo.wordpress.com/2013/03/12/bertrand-russells-inductivist-turkey/
6. https://www.reuters.com/article/us-britain-hawking-idUSTRE6811FN20100902 (accessed on
08/01/22)
7. https://en.wikipedia.org/wiki/Next_Generation_Air_Transportation_System#History
8. https://www.standishgroup.com/sample_research_files/CHAOSReport2015-Final.pdf
9. https://www.lexico.com/definition/science
Chapter 4
The Development and Application
of Knowledge

4.1 The Development of Knowledge—Principles

and Limitations

Although the development of knowledge is governed by general principles which I

explained above, it is useful to distinguish between fields of knowledge according to
their subject matter, which is defined by their fundamental concepts and relation-
ships and the body of knowledge developed on their basis.
Mathematics and logic are the most fundamental fields of knowledge as they
provide models and theoretical tools to represent the world and its phenomena. They
are necessary for every other field of knowledge, insofar as its specialized knowl-
edge can, of course, be formalized and studied as mathematical relations.
Following mathematics, I consider key fields of knowledge to be physics, infor-
matics, and biology, as they have separate and non-overlapping subject matters.
Physics is concerned with the study of matter/energy phenomena in space-time.
Informatics is concerned with the transformation of information—what we can
compute and how to go about computing it. It is not a field of mathematics, but it is
based on mathematical models. Additionally, it has an important technical compo-
nent related to building computers.
Biology is concerned with biological phenomena involving intertwined physico-
chemical and computational processes.
The other fields of knowledge can be considered composite since, apart from
specialized knowledge, they also use knowledge from these key fields. For example,
economics is a composite field of knowledge because it deals with relationships
between natural and monetary resources managed by people. The main problem is
how to best use resources in accordance with specific goals in order to meet
individual and social needs. The study of economic systems combines domain-
specific knowledge and more general knowledge from various fields such as psy-
chology, sociology, automatic control, and systems theory.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 31
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_4
32 4 The Development and Application of Knowledge

All fields of knowledge, regardless of their subject matter, have a common

approach characterized by the following three principles:
1. Modeling, i.e., using models that are representations of phenomena and systems
expressed through natural or artificial languages.
2. Stratification of modeled reality through an appropriate abstraction hierarchy
highlighting key properties to be studied for each modeling layer.
3. Modularity of the models used in each layer to break down the complexity of the
modeled reality and understand it as a synthesis of components.

4.1.1 Modeling—The Role of Language

Modeling is the process of representing a phenomenon or a system under study. It is

done using either a natural language or an artificial language, usually based on
mathematics, such as systems of equations, algorithms, or logic. As a rule, natural
languages allow for freedom of expression, as they are rich in concepts and struc-
tures. Conversely, artificial languages are built with a limited number of well-defined
basic concepts and structures. Their advantage over natural languages is the analyt-
ical ability of mathematics. Generally, expressiveness and analyzability are inversely
proportional: the more expressive a language is, the lower its analyzability.

4.1.1.1 Natural Languages

Ludwig Wittgenstein considers that “the limits of my language are the limits of my
world,” i.e., our language sets the limits of our understanding of the world [1].
Concepts and language, as they took shape empirically, are the first models of the
material and mental world. As for how they emerged through evolution and became
vehicles for cognition is another story, about which we can but speculate. It seems
that humans grasped physical numbers and the basic operations between them at a
relatively early age—as seen, for example, in the Sumerian astronomical tables.
Humans also tried to use myths to explain natural phenomena or to codify certain
ethical rules quite early. First, we created a kind of “technology” and “know-how,”
enabling us to solve a number of problems empirically and systematically. For
example, we were able to build wheels, light fires, measure quantities and areas,
cultivate land, and reproduce plants.
The first attempt to rationally interpret reality with what is called theory took
place in Greece. Pre-Socratic philosophers attempted the reduction of reality into a
few constitutive elements, seeking explanations based on laws. We thus saw the
advent of mathematics and the study of the properties of numbers with the Pythag-
oreans, Democritus’ atomic theory, Zeno’s paradoxes of the relationship between the
continuous and the discrete, Plato’s ideas, and Aristotle’s ingenious taxonomies.
These were followed by major steps in logic by the Sophists, the axiomatization of
4.1 The Development of Knowledge—Principles and Limitations 33

geometry, and the development of astronomy, which led to the heliocentric model
and the proof of the Earth’s sphericity.
The Renaissance saw the second major wave in the further development and
formalization of knowledge. Progress in mathematics was closely linked with the
development of mechanics, optics, and chemistry.
It is awe-inspiring that the human mind discovered simple mathematical and
geometric concepts by making powerful abstractions. We thus invented the concept
of a straight line, a circle, or a right angle. The physical world is rugged, it has hardly
any right corners and very few perfectly circular shapes. The human mind invented
the main abstract concepts underlying geometry, arithmetic, and continuous math-
ematics, which have enabled us, to some extent, to grasp the essence of physical
phenomena, and to achieve predictability.
The long trajectory of the intellect toward understanding the world is character-
ized by abstraction. It starts with the creation of names, where the mind carries out
generalization—it gives the name “stone” to a set of objects in which it discerns
common characteristics, despite differences in shape, color, texture, etc. This is the
foremost act for the mental interpretation of the world.
Generalization creates relationships of equivalence. This allowed for the creation
of categories of objects, leading to a first hierarchy of concepts. The emergence of
adjectives or adjectival modifiers plays an important role here. In the ancient Greek
language, one category of fruit was karya, with a hard shell, and another was mela
(with a soft rind). There are various kinds of karya with modifiers such as Pontic
(hazelnuts) or Persian (peaches), and various kinds of mela, such as Cydonian
(quinces) or Damascene (plums).
The advent of verbs and adverbs is the first sign of understanding change and
time. The basis for the invention of verbs is the construction of relationships linking
words, cause-and-effect relationships, or temporal relationships. First, there were
many irregular types, which abound in Homer’s poems. Over the course of its
evolution, language discovered laws of economy, systematizing the construction
of tenses and conjugations.
Other animals also possess simple languages to specify objects, actions, and
situations. However, the appearance of abstract concepts is a huge leap that makes
humans stand out. Much has been written about this, particularly thanks to the
analysis of the etymological provenance of concepts in old languages such as
Greek or Hebrew. Humanity owes much to ancient Greek thought, which
bequeathed the vast majority of philosophical, scientific, technical, and mathemat-
ical concepts.
The emergence of scientific concepts and theories is based on the power of
abstraction. These include the discovery of the properties of time and space in
order to understand natural phenomena. In physics, fundamental abstractions are
those concerning matter and energy. By means of the “game” of successive abstrac-
tions, we come to understand the physical world through laws that express relation-
ships between these two concepts that are characterized by measurable quantities.
Additionally, important was the use of money for transactions and, much later,
understanding the economy through the capital–labor relationship.
34 4 The Development and Application of Knowledge

Finally, other fundamental concepts are those of information and knowledge, the
importance of which we only grasped in the twentieth century. Theories of compu-
tation allow for the mathematical study of languages.
We can view the development of scientific knowledge as an attempt to devise
new languages rooted in mathematics with strong empirical foundation.

4.1.1.2 Formalized Languages—Theories

As we have said, mathematics and logic allow us to formalize knowledge in other

fields. When a relationship is expressed in mathematical terms, for example, as an
equation, it is strictly and unambiguously defined. On the other hand, relations
expressed in natural languages may be open to multiple interpretations and are
generally ambiguous.
Every mathematical theory is based on a set of axioms and rules for the concepts
it formalizes. Axioms are characteristic properties of theoretical concepts and are
generally inspired by empirical knowledge. We thus have the axiom that for every
number n, the number n + 1 is different from n. Another axiom is that a proposition
can be either true or false, but not both.
The rules of a theory allow us to infer theorems through axioms. One character-
istic rule of propositional calculus is that of deduction: if a is true and if a implies b,
then b is true. This rule is followed in the well-known example: All men are mortal
and Socrates is a man, therefore Socrates is mortal.
Euclidean geometry is the first axiomatized theory proposed by Euclid of Alex-
andria in the fourth century BC. It thus demonstrated the power of formalization and
paved the way for logic and abstract mathematical thinking. Only five axioms and
three basic concepts are sufficient to formalize plane geometry.
One abstract concept that is very important for mathematics and their application
is that of infinity and, more generally, infinite quantities. It has been proven that there
are different types of infinity, and an order relation can be defined between them,
similar to the order between natural numbers.
How can we prove relationships between the infinite elements of a set? This
requires a different kind of logical rule: the rules of induction. These rules apply to
hierarchically structured sets and allow for generalization on the basis of the
following reasoning: “If a property is valid for the smallest elements of the hierarchy
and its validity for layer n implies its validity for layer n + 1, then the property is
valid for all the elements of the hierarchy.”
One classic example is the induction on the set of natural numbers with the
hierarchical relation n < n + 1, which we learn in high school. The general
application of induction to infinite sets requires the definition of an appropriate
hierarchical relationship, which generally depends on the type of property we wish
to prove. For example, if I wish to prove that a program terminates, I need to find a
function associating a positive quantity to program states and such that for all
program runs, this quantity steadily decreases.
4.1 The Development of Knowledge—Principles and Limitations 35

One consequence of Gödel’s theorems (see Sect. 3.4.2) is that there is no

algorithm that decides program termination. This result entails that discovery of
induction rules cannot be automated.
Contradiction arises in a theory when a proposition and its negation are true.
Contradictions violate a well-known principle of reasoning. A mathematical theory
must be “consistent,” i.e., it must not contain contradictions. It should also be
“complete,” i.e., every true proposition must be provable using axioms and inference
rules.
In the early twentieth century, the German mathematician David Hilbert proposed
a research program aimed at proving that mathematics is consistent and complete.
Furthermore, he aimed to prove that mathematics is decidable, i.e., there is an
algorithm that decides the truth or the falsity of any proposition in the system.
I have already mentioned Gödel’s theorem, which states that in powerful enough
consistent theories there are undecidable propositions (see Sect. 3.4.2). Gödel also
showed that theories of arithmetic (integers with the operations of addition and
multiplication) cannot be proven to be consistent—that is, free of contradictions—
within the theory, even when they are consistent.
One way of violating the consistency of a theory is to allow self-reference, i.e.,
propositions that refer to themselves. The “liar” paradox, which is attributed to the
philosopher Eubulides of Miletus, who lived in the fourth century BC, has been
known since antiquity: “A man says that he is lying. Is it what he says true or false?”.
By applying the rules of logic in this case, we reach the well-known “vicious cycle.”
The proposition implies its negation, which implies the proposition.
This paradox remained unexplained for 2000 years, until Bertrand Russell for-
mulated the paradox in set theory in the early twentieth century, proving that the
standard model of set theory cannot prove its consistency.
One way of understanding the practical significance of these results is that any
consistent theory encompassing arithmetic is “open” in the sense that there are
propositions that cannot be proven or disproven in the theory.
A scientific theory condenses and formalizes scientific knowledge derived from
the process described in Sect. 3.4.3. It takes its models from a mathematical theory to
which special axioms are added that reflect the fundamental laws governing the
phenomena we are studying.
For example, in Newton’s theory, the mathematical theory is differential calculus,
where the variables represent the key concepts that characterize physical objects and
their state in space-time, for example, mass and shape, coordinates, velocity, and
acceleration. The principle of conservation of energy and the law of universal
gravitation are added to the theory as axioms. Of course, a scientific theory must
be consistent in terms of the mathematical concepts it formalizes, i.e., it must not
contain logical contradictions. However, this is not enough. It must also be valid, i.e.,
no theorem should be falsified by observations.
I explained that our ability to understand phenomena also depends on the arsenal
of models available to us. We have often devised new concepts and models, which
while contrary to “common sense,” simplify the study of phenomena. Examples
36 4 The Development and Application of Knowledge

include the introduction of the concept of infinity to mathematics or the use of the
Dirac delta function in physics.

4.1.2 Stratification—Abstraction Hierarchy

The physical world has breadth and depth. Phenomena can be studied at scales
ranging from 10 35 m, which is the order of the Planck length, the smallest
measurable length, up to 1026 m, which is the size of the observable universe. In
order to understand these from the very small to the very large and in order to
overcome the complexity of this 1061 change of scale, we study the world at various
layers of abstraction. Please note that is just a methodological simplification and it
does not at all mean that reality is stratified! In contrast, I believe that reality is a
single whole—however, this is an ontological question.
First, let me explain what abstraction is, as it must not be confused with
ambiguity and indeterminacy. Abstraction is a holistic approach to breaking down
complexity, highlighting essential characteristics of a phenomenon for each level of
observation. Thus, when I observe the world with the naked eye and I am studying
physical phenomena or building artifacts, I can forget that the matter is composed of
particles; I can even consider that a solid, material body is a point where its entire
mass is concentrated. The use of abstraction has proved essential in order to tame
complexity.
We use model hierarchies in every field of knowledge, where each layer is
associated with the superior layers through an appropriate abstraction relationship.
As we climb up the abstraction hierarchy, the scale of observation increases.
Figure 4.1 presents an indicative hierarchical stratification for physics, informat-
ics, and biology. Readers may be more familiar with that of physics, which ranges
from the particle world to the universe. It goes through atoms, molecules, and
theories of solids, liquids, and gases, which are useful to study complex electro-
mechanical systems that make up artifacts.
Similar stratification can be carried out for computing systems, which in their
physical form are integrated circuits implementing the logical gates and memory of
computers. The upper strata are networked systems that make up the cyber-world.
An obvious requirement when stratifying knowledge is its unification, in other
words, to define relationships between the models at each layer so as to relate the
results and achieve a global understanding and knowledge of the field being con-
sidered. This would lead to what certain physicists call a “Theory of Everything.”
Such a unification of knowledge has not yet taken place in any of the key fields noted
above. In the case of physics, for example, the general theory of relativity would
have to be linked to quantum mechanics, which is known to be an open problem.
The integration of stratified knowledge comes up against insurmountable prob-
lems that will be analyzed in what follows.
4.1 The Development of Knowledge—Principles and Limitations 37

Fig. 4.1 Hierarchical stratification for the three key fields of knowledge

4.1.3 Modularity—The “Atomic Hypothesis”

In each layer of an abstraction hierarchy, to break down the complexity of the

observed reality, humans make the very useful hypothesis that the world is built
from components, building blocks, composed based on laws that we try to discover.
Democritus was the first to have this idea and made a very interesting argument in
favor of the “atomic hypothesis” concerning nature. In other words, he tried to prove,
through logical reasoning, that the physical world is not continuous but discrete.
Of course, the ontological question “is nature discrete or continuous?”, which
cannot be approached rationally, is less important. What is important is the fact that
the “atomic hypothesis” allows humans to break down complexity and develop
scientific knowledge, i.e., to understand phenomena rationally.
The development of physics and chemistry is the triumphant proof of the effec-
tiveness of such a modularity assumption. Matter is made up of a small set of
elementary particles that, through successive combinations, result in the complex
universe. We construct buildings by combining building materials such as bricks and
concrete blocks. We manufacture electrical circuits by connecting resistors, capac-
itors, and inductors. We build machines by assembling parts. In every case, theory
allows us to predict the behavior of the whole, by composing the behavior of its
components. The composition of components can also be defined in a mathematical
or systematic manner. Thus, the behavior of a circuit is defined by the behavior of its
components and the way they are connected.
38 4 The Development and Application of Knowledge

We should also note that natural languages evolved according to this building
principle: the meaning of a phrase is a synthesis of the meanings of its words.
As a methodological approach, which considers a complex system to be the
composition of a relatively small number of types of components, the modularity
assumption is based on the following three basic rules:
1. The whole is a composition of a finite number of types of components. For
example, an atom consists of electrons, protons, and neutrons; similarly, a
chemical compound is a combination of different kinds of atoms. A cell is a
synthesis of cell organelles. A phrase is made up of words.
2. We can learn the characteristics of each type of component by studying it
individually, just as we can separately examine the behavior of particles, atoms,
or the meaning of words.
3. Our knowledge of the whole can be inferred by synthesizing our knowledge of
the components that make it up and the place of the components in the structure of
the whole.
It must be noted that if, thanks to the second rule, we can learn the behavior of
each type of component individually and know how the whole is structured using
different types of components, then by applying the third rule we can know the
behavior of the whole.
For example, by knowing the behavior of the components of an electrical circuit
(voltage source, resistor, capacitor, and inductor) and the connections, we can study
the behavior of the circuit. Our knowledge regarding each type of component is
equations that associate current with voltage. The connections state how we can use
the equations of the components to determine a system of equations characterizing
the behavior of the circuit.
Note that in this example, the synthesis of global knowledge assumes the locality
of interactions between the components making up a system. Each component
interacts with components in its “neighborhood.” This assumption of the locality
of interactions does not appear to hold true in quantum physics or in biology, which
makes the relevant phenomena much more difficult to model and understand.
Furthermore, the application of the third rule requires that the behavior of a
component we have learned by studying it separately does not change when we
synthesize it with other components or, at the very least, that it changes in a
predictable way. Unfortunately, this hypothesis does not hold true for information
and biological systems, nor for natural languages. It is well known that the meaning
of the words of a phrase depends on context, and this makes it extremely difficult to
translate natural languages. Furthermore, informatics lacks satisfactory results that
allow us to know the properties of a software by synthesizing the properties of its
components.
This discussion shows that modularity that was so helpful in our understanding
the physical world does not carry out in fields of knowledge such as informatics and
biology. It applies even less to economics and social sciences, where systems are
composed of components, for example, individuals, whose behavior is difficult to
study in isolation without taking into account their position within the structure of
4.1 The Development of Knowledge—Principles and Limitations 39

the system. Human behavior depends so much on the context in which it takes place
that it is practically impossible to characterize.
In conclusion, I would point out that when a system consists of a large number of
components, but the behavior of each component remains unchanged, however
much the number of system components changes, then we can study the behavior
of the system using statistical methods. This is the case with the kinetic theory of
gases, which studies gases as ensembles of particles that obey the laws of Newtonian
mechanics.
However, the statistical approach can only be applied to “disorganized complex-
ity” phenomena involving a very large number of equivalent components, as the
economist Friedrich August von Hayek noticed [2]. It cannot be applied to phenom-
ena of “organized complexity” observed in biological, information, or economic
systems. This is because, as I explained above, the behavior of each component
changes dynamically depending on its position in the system structure. The study of
behavior and the control of systems of organized complexity are one of the most
important challenges in science.

4.1.4 Emergent Properties

We have explained how stratification allows for a piece-by-piece understanding of

phenomena. Each layer is characterized by a specific abstraction level with its own
basic concepts and laws. Going up the hierarchy, some concepts may not be applied,
while others emerge as important. Thus, stratification may introduce gaps in the
understanding of the phenomena under study.
I have already said that a very important problem in every field of knowledge is
the consolidation of knowledge between different levels of abstraction: can we
explain the properties of components at one layer from the properties of the layer
immediately below?
For example, can we know
1. the properties of water molecules from the properties of the oxygen and hydrogen
atoms and the rules governing their composition?
2. the properties of an information system from the properties of components of its
hardware, properties of the programs being executed and the initial memory
state?
3. the properties of our mental processes from the properties of the brain’s neural
systems?
These questions are of a similar nature and a satisfactory answer is most probably
unlikely. Without getting lost amidst technical explanations, I would say that there is
a problem of scale. When we move from one layer to a higher layer, new properties
“emerge” due to the interaction of a large number of components that form much
more complex behaviors as a whole. Of course, just saying that new properties
40 4 The Development and Application of Knowledge

“emerge” does not explain anything. We are simply acknowledging our inability to
consolidate the theories of a field of knowledge.
For instance, we know that water consists of a set of molecules and that each
molecule is a synthesis of oxygen and hydrogen atoms. Even if I could theoretically
determine the behavior of a molecule, there are properties that are the result of the
interaction of molecules. A 250 ml glass of water contains 8.36  1024 molecules.
The emergent properties of water are that it evaporates at 100  C, freezes at 0  C and
that volume changes according to temperature. Even if it were theoretically possible
to faithfully describe the dynamic behavior of molecules and their interactions,
modeling them would be extremely complex and practically impossible to resolve.
Similar considerations apply to both other examples. When we view a computer
as an electrical circuit described by systems of differential equations, it is practically
impossible to go up the abstraction hierarchy and achieve a match between the state
of the circuits and the execution of the software.
Matching the properties of the brain with the properties of our mental functions
seems to be even harder. Unfortunately, there are scientists who are “over-optimis-
tic” about the possibility of overcoming these difficulties. For example, the Human
Brain Project, which received over 1 billion euros in funding from the European
Union [3], promised that by studying and simulating detailed models of the brain, it
is possible to understand phenomena concerning consciousness. Squandering money
on infeasible but “catchy” is a common feature of major research programs.
Our inability to explain the emergence of properties is discussed by Philip
Anderson in his article entitled “More Is Different” [4], of which I quote a passage
below, because it very clearly explains that our ability to reduce the whole into
simple laws does not imply that, starting from those simple laws, we can once again
find all the properties of the whole.
The main fallacy in this kind of thinking is that the reductionist hypothesis does not by any
means imply a ‘constructionist’ one: The ability to reduce everything to simple fundamental
laws does not imply the ability to start from those laws and reconstruct the universe. In fact,
the more the elementary particle physicists tell us about the nature of the fundamental laws,
the less relevance they seem to have to the very real problems of the rest of science, much
less to those of society.

4.2 Issues Related to Knowledge Application

4.2.1 The Limits of Scientificity

As I said, many believe that there is a clear and absolute distinction between
scientific and non-scientific knowledge. We must not, of course, overlook the
differentiation in the validity of knowledge, but an absolute distinction between
scientific and non-scientific is completely arbitrary and dangerous.
I explained that the knowledge is valid by degrees. Even in the natural sciences,
for example, astronomy, the possibility of experimental validation and repeatability
4.2 Issues Related to Knowledge Application 41

is limited. We must also distinguish between experimentally confirmed relationships

(laws) and the interpretations we normally give for the sake of generality and
consolidation. Thus, in the past we believed that space was full of “aether” or in
the existence of “phlogiston,” a fire-like element contained within combustibles and
released during combustion.
What often strikes me is the confidence with which renowned scientists discuss
grand theories. They make a clear effort to impress the general public in order to gain
publicity, popularity and, consequently, financial support for their research.
Every time a theory is proposed, the criteria that permit its refutation must also be
given.
There are theories considered to be categorical truths that, when examined
according to logical and methodological criteria, give rise to questions and doubt.
Such theories are those trying to explain what happened in the past based on present-
day observations, such as cosmological theories and theories on the evolution of
species. The formulation of these theories raises a particular logical problem.
While, in order to discover physical laws, we study a cause-and-effect relation-
ship in the here and now, in the case at hand, we do know the outcome, but we have
to guess the root causes. We must perform a logical operation called abduction,
which, unlike deduction that proceeds from the hypothesis (cause) to the conclusion
(effect), seeks to find the “most likely cause.” The difference with deduction is that
many hypotheses can logically lead to the same conclusion.
Some people liken this logical problem to the problem of a detective who arrives
at a crime scene and has to guess murder scenarios only from on-the-spot observa-
tions, without interviewing witnesses.
Cosmological theories are based on certain hypotheses, which we consider valid
as long as they have not been refuted. Such hypotheses are that the universe is
isotropic and that the physical laws that we observe apply uniformly with the same
values of universal constants and, in particular, that there are no faster-than-light
communications.
Generalizing theories validated in our “neighborhood” to apply to the entire
universe has a certain amount of arbitrariness. Perhaps we should be slightly more
careful and modest when trying to solve cosmological problems, particularly when
the prevailing cosmological theory has significant weaknesses. How pleased can one
be with a theory, which posits that only 5% of the universe contains matter and
energy such as that found in our solar system, the remaining 95% being dark matter
(27%) and dark energy (68%)? The entire Big Bang narrative leaves logical ques-
tions unanswered, and I do not know if and when these questions might be given a
satisfactory answer.
However, even Darwin’s theory of evolution, which everyone now accepts as an
indisputable truth, has certain logical gaps, which must not be suppressed. First, I
would like to state that I do not question the idea of evolution, which is clearly
confirmed by observations. However, what needs to be explained scientifically as a
possibility is how a single-cell organism can evolve into multicellular organism with
eyes, ears, and other organs. How, by applying the basic idea that through a
succession of random changes and choices of the fittest (natural selection), we can
42 4 The Development and Application of Knowledge

reach multicellular structured organisms such as humans. I believe that Darwin’s

idea of natural selection explains the adaptability that can be achieved through
learning processes. Technically speaking, it is a process of changing parameters in
order to optimize a criterion, which in this case, concerns the viability of the
organism. However, what remains technically unexplained is the creation of com-
plex organisms through this “game” of random chance and consequent adaptations
to the external environment.
Another way of exploring the problem is to think of it using an analogy:
comparing the construction of an artificial system to the development of a biological
system. Those who design systems or write software will understand precisely what I
am talking about. Optimizing a structured system already in existence is one problem
and constructing a system by composing components that add new functionality and
cooperate harmoniously is another altogether. Everyone understands that it is not
possible for a radio to become a television set by applying a mechanism such as
natural selection, no matter how we imagine interaction with the environment.
From this analogy, we can reasonably conclude that biological systems differ
radically from artificial ones, and that no satisfactory explanation will be found as
long as we limit ourselves to the organism/environment “game” without taking the
very properties of organic matter into account. One such property overlooked by
Darwin’s theory is that organic matter, and only organic matter, is not only able to
learn, which largely explains adaptability, but can also self-organize. In my view,
these properties that we have to discover make the game of evolution less random.
Perhaps studying genes alone is not enough in order to understand building
mechanisms and the emergence of new functions. We must try to describe the
“grammar” of the construction of these biological systems, i.e., structuring principles
of components at the various layers of the bio-hierarchy. As is the case with other
fields of knowledge, the search for theories that provide a holistic “bottom-up”
explanation risks proving exceptionally fruitless and disappointing.

4.2.2 Scientism

One implicit hypothesis of physics that, to some extent, can be confirmed experi-
mentally is that each factor that affects a phenomenon is observable and measurable.
This does not apply to other fields of knowledge. For example, social phenomena
cannot be understood by examining relationships between measurable quantities.
Thus, the human factor is frequently ignored in economic models, simply because its
behavior cannot be rigorously modeled.
The successes of physical sciences and their technical applications, particularly
during the nineteenth and twentieth centuries, gave rise to a philosophical move-
ment, which considers that only mathematically formulated knowledge is worthy of
trust and recognition. This led to the unchecked use of mathematical models in many
fields of knowledge with mediocre to completely disastrous results.
4.2 Issues Related to Knowledge Application 43

The trend named scientism in the mid-twentieth century concerns the uncritical
application of models and approaches from the exact sciences. A typical example of
scientism is the failure of economists to formulate successful policies because of
their propensity to emulate the physical sciences as closely as possible. Thus, the
correlation between aggregate demand and total employment is the only one for
which we have quantitative data and will therefore be accepted as the only causal
connection that exists [2]. A theory based on measurable data can be considered
more plausible than another one that provides certain well-founded explanation,
which, however, one can challenge due to the absence of measurements.
We often liken the search for knowledge to the behavior of the drunkard who, as
the old joke goes [5], has lost his car keys and is searching for them under a
lamppost. A policeman sees him and helps him look together. After a short while,
the policeman asks the drunkard whether he is sure he lost the keys there, and the
drunkard replies that, no, he lost them in the park. The astounded policeman then
asks why he is searching for them under the lamppost, and the drunkard answers
“because that’s where the light is.”
This happens more often than you would think in the search for knowledge.
Researchers prefer to write publications about beautiful theories that are far removed
from reality rather than racking their brains over difficult practical problems. Few
take the trouble of judging the validity of their hypotheses and the applicability of
their results. This is a phenomenon that I have also observed in informatics, where
beautiful theories propose models that over-simplify reality.
It is not surprising that scientists awarded the Nobel Prize in Economics are
allegedly mediocre at business. Econometricians reduce problems to equations and
ignore the human factor, which is also the main parameter in an economic game.
They ignore the qualitative. The exact same economic policy would produce very
different results when applied in two different countries, for example, Germany and
Greece.
There are serious reasons to be concerned about the dangers of blithely adopting
approaches that are only ostensibly scientific. The progress achieved in the physical
sciences over the last decades has exceeded every expectation and any discussion
about the existence of limits could be considered suspicious.
Public opinion fails to understand that some ideas promoted under the guise of
science may be at best empty, if not misleading, and therefore dangerous. There are
countless examples in every field of knowledge.
I remember being an informatics student in 1972, when the Club of Rome
published a famous text, in the name of an ill-defined “systems science,” entitled
The Limits to Growth [6], which predicted a pessimistic future for humanity, an
immediate shortage of energy resources and raw materials, and their impact on
industrial development and the environment. The text was published as a book,
sold 30 million copies, was translated into more than 30 languages and remains the
top best-seller of books on the environment. I remember that, at the time, they
lavishly financed research projects for the development of software simulating
economic systems. The conclusions of the text were widely publicized by the
media and seriously influenced economic policies around the world; conversely,
44 4 The Development and Application of Knowledge

very little was heard of the radical criticism rightly voiced by authoritative scientists
and analysts. Of course, when we now examine what happened at the time, it is clear
that the supposed analysis capabilities of then-mainframes were used for economic
and political purposes.
It is imperative that we protect ourselves from scientism, which derives its
authority from methodologically unsubstantiated analogies to the physical sciences.
The risk of controlling social processes in the name of supposedly science-based
predictability—especially at present with the use of artificial intelligence—is
absolutely real.

4.2.3 Experts—The Mystification of Expertise

Scientism often goes hand-in-hand with a dangerous trend of mystifying expertise,

which holds that all serious matters must be judged and ultimately decided upon by
experts. This mystification is compounded by the management of knowledge by
computers.
Today, there are experts on every issue, ready to advise and guide us. They range
from those who would advise us on personal problems to those who propose ready-
for-use solutions to major social and national issues, for example, how to overcome
the economic crisis, how to reduce inequality, and how to address the refugee issue.
This is not to say that we do not need experts to shed light on technical aspects of
the complex problems facing societies. However, the solution to every major
problem also has a political and moral dimension. When the time comes for choices,
public opinion, once properly informed, must be able to take part in making crucial
decisions.
I believe that it is possible to explain to the public in a comprehensible way what
is at stake in a decision involving risk and what its consequences are. What is
necessary is an effort for popularization, for appropriate awareness raising and
preparation through education and public debate. Access to knowledge and the
development of a critical attitude in the many is essential in order to enable them
to participate actively and knowingly in processes and decisions that determine their
future.
If crucial decisions are made without transparency by an exclusive circle of
experts, advisers, and politicians, this poses a danger to democracy and liberties.
It is interesting to compare how the coronavirus crisis has been managed in
different countries. At first, the experts did not agree with each other on the
magnitude of the risk and the precautions that had to be taken. However, what
made a difference was not so much the knowledge of experts as much as decisions of
a political and moral nature. Certain countries made it a priority to save lives over
other criteria such as not disrupting the economic life of the country. By this, I mean
that experts define a framework of options and the state makes decisions based on
value criteria (see Sect. 8.2.2.2—Risk Management Principles and Their
Implementation).
4.2 Issues Related to Knowledge Application 45

Knowledge has long been the preserve of the few. Since antiquity it was used to
prop up the powerful, who could buy knowledge and use it as they saw fit. I must
point out that education is set up in such a way so as to favor those who memorize,
affording little acknowledgment to those possessed of creative and inventive minds.
It does not foster curiosity and creativity. It favors intellectual indolence and
conformism. At the same time, I am not certain that the possibilities afforded by
the Internet to access knowledge have been exploited to the full by providing
appropriate services and incentives.
Experts today play an important role in shaping public opinion and directly
influence political choices. Acting either as independent professionals or through
famed consulting firms and think tanks, they constitute an institutional ecosystem
that affects economic and technological development. They usually arrive with
in-depth analyses in hand—replete with countless references, graphs, and
reports—either to endorse certain political choices or to throw up a smoke screen
and muddy the waters.
In many countries, experts lend the authority of expertise to disastrous policies
and serve as a necessary complement to the status quo. As a Greek, I experienced
first-hand the terrible economic crisis which broke out in Greece in 2007 and which,
under the advice of famous academic and institutional experts, led to the application
of a series of reforms and austerity measures that proved to be disastrous for the
country’s economy. It is now more than clear that the country underwent a harsh
treatment that resulted in dramatic impoverishment and loss of income and property
that compromised the country’s independence and mortgaged its future.
In addition to its validating role, the experts’ ecosystem may also be parasitic,
depending on the environment. I have read studies by experts proposing changes for
Greece that were hollow, if not malicious. I have seen committees of experts
invested in writing reports, spending countless hours and funds to produce findings
that, in the best of cases, end up in the waste bin. Of course, even when they reach
specific conclusions and proposals, these prove convenient for those who
commissioned them.
The mystification of expertise and the alienation of public opinion from knowl-
edge, which is evolving rapidly and is becoming increasingly difficult to understand,
is a danger for modern societies.

4.2.4 Researchers, Research, and Innovation

I will end this chapter on the development of knowledge by saying a few things
about organized research and researchers.
46 4 The Development and Application of Knowledge

4.2.4.1 Research and Researchers

Scientific research is currently aimed at the systematic production of knowledge. It is

essential for the development of a modern, innovative economy. It is conducted at
universities and research institutions, as well as laboratories of large companies.
The profession of researcher differs from other professions, as it requires strong
motivation and dedication, as well as a certain talent for creativity. What I found
wonderful in this profession is the sense of freedom, a prerequisite for exploring new
pathways for creation. However, managing this freedom while possessed of a sense
of responsibility is not a simple problem. I have seen researchers lose themselves in
choices leading to impasses, as well as researchers who prefer safety in taking beaten
paths, which are free of risk but where there is little chance of breakthrough results.
A professor used to tell me that “you will not invent the light bulb by trying to
improve on candle-making.”
Throughout my life, I tried to carve out my own path by following a vision of my
own. At times, I achieved this, and at others, I did not. I have tried to keep myself
outside mainstream thinking, with which I did not agree, and that has had and still
entails a certain cost.
There are currently large communities of researchers, each working on specific
issues, who share a common concern and who try to push the boundaries of
knowledge as far as possible. Research is less free than it was 60 years ago. It is
driven by large programs that are generously funded by states and businesses and
aim at what we call scientific and technical challenges. The programs are multi-
annual and each absorbs significant research funds. That is why it is often discussed
whether and to what extent the funds spent on a program will deliver the expected
results within a reasonable period of time. Without elaborating on this very impor-
tant topic, I would stress that there must be strong social control over the priorities
and directions of research.
I raise the issue in order to underline the difference between a well-defined
scientific and technical challenge, on the one hand, and a vision, on the other. A
challenge has as its objective to overcome an obstacle in the development of
knowledge, and this objective has the following characteristics:
1. The objective is relatively well defined and concerns: either achieving a specific
goal such as proving a difficult mathematical result and developing a vaccine or
seeking a framework that consolidates existing results and proposes a new
approach to the development of knowledge, for example, the unification of
theories in physics.
2. The objective is realistic in the sense that it takes into account the state of the art in
the field of knowledge in question, as well as theoretical and practical limitations,
for example, complexity and experimental evidence, respectively.
3. Achieving this objective will constitute a breakthrough in the development of
knowledge. In other words, it will bring about a significant qualitative change in
knowledge that is different from everyday gradual progress.
4.2 Issues Related to Knowledge Application 47

With this characterization, I wish to make a clear distinction between challenges

and visions. A vision also represents a long-term ambitious goal, but it does not meet
any of the above criteria. For example, one such vision was the “Strategic Defense
Initiative,” the so-called Star Wars programme, launched in 1983 by US President
Ronald Reagan [7]. It was clear from the outset that the objectives of this program
were not achievable and I think no expert ever believed otherwise. Although the
program failed to achieve its original objectives, it mobilized investments in the
defense sector that may have yielded significant “collateral” results.
Today, the AI vision mobilizes the massive investment and the enthusiastic
involvement of big tech companies. Nonetheless, there is a lot of confusion about
what intelligence really is. The media and big tech companies, with large-scale
grandstanding projects, spread fallacious opinions suggesting that human-level AI
is only a few years away. The excitement and buzz about the advent of a new
intelligence era are amplified by the predictions of various “experts” about a future
inevitably dominated by machines. This does not allow a sober, informed, and
reasoned debate to ponder the risks of the AI revolution that most people tend to
accept as a fatality we cannot escape.
The problem with visions is that, beneath visible high and fancy objectives, they
may hide other objectives, unspoken and less noble.
What are the chances that the work of CERN, the European Organization for
Nuclear Research, which has an annual operating budget of one billion euros, could
achieve the goals it has set? These include pushing “the frontiers of science and
technology for the benefit of all” and helping to “discover what the universe is made
of and how it works” [8]. It is difficult to imagine how adding to the already long list
of discovered particles, new particles can contribute to achieve these grandiose
goals.
What needs do space programs serve? Why would we want to colonize other
planets? And even if this is possible on a small scale, the costs would be unbearable
and the benefit small. A manned journey to Mars involves enormous technical
difficulties—and yet the media often talk of colonizing this planet as if it were
possible in the near future.
“Visionaries” often invoke “progress in scientific knowledge” regarding all the
above. However, the issue at hand is one of the priorities and hierarchy of objectives.
There are more urgent challenges raised by environmental and epidemiological
problems at the global level that do not cause the same kind of enthusiasm. It
would therefore be good if the compelling visions asserted by large organizations
and maintained by powerful lobbies with the support of the media were analyzed in
depth according to the criteria mentioned above.
Research is costly today. It makes no sense to squander huge sum of money
simply to satisfy some curiosity or to achieve “progress in knowledge” in a general,
vague sense. Research must contribute to the well-intended progress of humanity
and have clear objectives that do not run counter to the common good.
48 4 The Development and Application of Knowledge

4.2.4.2 Research and Innovation

The emergence of innovation as a driving force of economies is a relatively recent

phenomenon. The research and technology model that developed after World War II
changed radically in the 1990s, mainly for two reasons.
First, there was an acceleration of the innovation cycle, where basic and applied
research worked in tandem, particularly under the pressure of competition and the
market. Second, the development of basic research became a luxury, feasible only
for very large companies.
Today, businesses are less vertically integrated. In order to minimize costs and
risk, and in cooperation with research institutions, they develop research and tech-
nology of common interest that they tailor to meet their own needs. In such partner-
ships, venture capital plays an important role in exploiting research results by
creating start-ups.
Collaboration between businesses and universities has radically changed the
structure of research at universities and the criteria for evaluating researchers. We
have transitioned from small theoretical research teams to large laboratories, where
hundreds of researchers collaborate on ambitious research programs.
I remember, as a child, picturing a researcher, wearing a white coat and thick
glasses, hunched over a laboratory desk, working night and day to come up with a
magical formula. Today, we are far removed from this model of knowledge produc-
tion. We need significant funds and infrastructure, as well as critical mass and
excellence.
Critical mass is essential because research is closely linked to technology and,
therefore, there needs to be cooperation between basic and applied research. Fur-
thermore, research is often interdisciplinary and must be linked to relevant fields of
knowledge.
We also need excellence, which means the capacity to generate knowledge and
apply it to innovative ideas. The creativity of a talented researcher cannot be replaced
either by financial resources or by masses of inexperienced researchers.
The production of innovation is based on the creation of innovation ecosystems,
which are the result of synergy between three factors: (1) research institutions,
(2) industrial enterprises, and (3) start-ups. The most illustrious examples are to be
found around major technology universities in the USA such as Stanford, Berkeley,
MIT, and CMU.
An innovation ecosystem brings together significant human resources from pools
of experts, scientists, engineers, and managers and is characterized by a special
culture of creativity. It is attractive due to the cultivation of a special lifestyle—see
the case of Silicon Valley. The state intervenes by providing incentives and plan-
ning, ensuring access to foreign technologies, special tax regimes, and copyright
protection. Financial institutions—venture capital funds and banks—support start-
ups. Of course, everyone works together in order to produce innovative products and
services.
References 49

The admirable complementarity of the roles in the synergy between the three
actors in an innovation ecosystem must be emphasized. Research institutions carry
out basic research and work with industrial enterprises to implement it. They thus
secure research funds from their industrial partners and enjoy contact with the new
problems brought about by technological development. Start-ups are often founded
by researchers and inventors. They have the advantage of flexibility and efficiency
compared to large industrial enterprises.
Innovation is not the privilege of superpowers and large countries such as the
USA, Japan, and Germany. Small countries, such as Israel, Switzerland, and the
Nordic countries, are among the top in the innovation arena. Each of these countries
has developed innovation ecosystems and robust research structures based on the
advantages and strengths gained over the last decades. The most striking cases are
certainly Israel, a technological superpower, as well as Switzerland, which has a very
“aggressive” policy for the purchase of patents and rights.
In conclusion, I would like to emphasize the importance of the human factor for
building an innovation-driven economy. Lack of vision cannot be offset by money.
Policies should support creative individual initiative, favor meritocracy, and fight
corruption. They should implement measures fostering entrepreneurship and
attracting innovative businesses. They should reform research institutions to help
them achieve critical mass and excellence and provide incentives for connection to
the real economy.
There are countries with huge scientific potential that are doomed to fail simply
because their policies privilege economic and geographic determinism and ignore
how much the human factor matters in the innovation race.

References

1. https://en.wikiquote.org/wiki/Ludwig_Wittgenstein
2. https://www.nobelprize.org/prizes/economic-sciences/1974/hayek/lecture/
3. https://en.wikipedia.org/w https://en.wikipedia.org/wiki/Human_Brain_Project iki/Human_
Brain_Project
4. https://www.tkm.kit.edu/downloads/TKM1_2011_more_is_different_PWA.pdf
5. https://en.wikipedia.org/wiki/Streetlight_effect
6. https://en.wikipedia.org/wiki/The_Limits_to_Growth
7. https://en.wikipedia.org/wiki/Strategic_Defense_Initiative
8. https://home.cern/about/who-we-are/our-mission (accessed on 08/01/22)
 Part II
Computing, Knowledge and Intelligence

We analyze the relationship between computational processes and physical phe-

nomena, as well as the production and application of knowledge by machines and
humans.
• We investigate relationships between informatics and physics and discuss how
each of these fields of knowledge can enrich the other, comparing basic concepts
and models.
• We attempt to compare artificial and natural intelligence, trying to answer the
question of whether and to what extent computers, as they stand today, can
approach human mental functions.
• We present autonomous systems that are called upon to replace humans in
complex operations by realizing the vision of strong AI and discuss the risks—
real or hypothetical—of the reckless use of these systems.
Chapter 5
Physical Phenomena and Computational
Processes

5.1 Scientific Knowledge and Computing

The models of the physical world and those of informatics share a common feature
that can serve as a basis for comparison between them: they are dynamic systems
with states and actions that constitute transitions from one state to another. However,
there are also very important differences in the way we develop knowledge and the
way we formulate and interpret it. One difference already explained is that scientific
knowledge is empirical, while our knowledge of computing systems is a priori, at
least in terms of theory of computation.
In what follows, I will explain the similarities and differences between computa-
tional processes and physical phenomena. Of particular interest is the concept of
natural computers leveraging physical processes since it adopts a completely differ-
ent computation model that is not subject to the intrinsic limitations of the conven-
tional algorithmic model.

5.1.1 Computational Processes and the Experimental Method

The models of computation, as defined by A. M. Turing, are discrete, as they are

based on arithmetic. Computation is the process where a machine executes an
algorithm by performing a finite number of actions (steps) in order to calculate the
value of a function for given values of its arguments. When the process terminates,
the state of the machine provides the result.
Machines are modeled in terms of a mathematical relationship that determines
how the machine transitions from one state to another. A state consists of the values
of the data that the machine transforms when executing the algorithm. I have already
explained that the result of the computation is independent of what is happening in
space-time. One could define a concept of “logical time” as the number of steps a

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 53
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_5
54 5 Physical Phenomena and Computational Processes

machine has performed after it started executing an algorithm. However, this process
is sequential, and the concept of logical time is independent from physical time.
I have pointed out that our knowledge of computing machines is based on
mathematics and therefore different from our knowledge of physical phenomena.
In order to understand the behavior of a machine, the experimental method applied to
develop scientific knowledge is precarious.
Let us assume that we have a program and want to check if it can compute the
function f: x ! x2. In other words, when I input the number x into the program, it will
output the square of x as the result of its execution. Let us perform an experiment
where for values of x ¼ 0, 1, 2, 3. . ., I receive the result of the computation f(x) ¼ 0,
1, 4, 9. . . For what number of correct experimental data can I reasonably reach the
generalization that the program actually computes the square of x? Why does the
inductive method of generalizing observations, which applies to physical phenom-
ena (see Sect. 3.4.3.2—On the Nature of Scientific Knowledge), makes no sense for
algorithms and, therefore, for the programs that describe them?
This is because simple physical phenomena are robust, in the sense that small
changes have isometric effects. This does not apply to algorithms due to their
discrete nature. Changing slightly in memory can change their behavior radically.
However, since algorithms are defined as mathematical models, we can—at least
in theory—achieve stronger guarantees through mathematical verification, that is to
say, to prove a theorem based on logical analysis of the algorithm’s code, that for
each value of x it will compute x2.
Thus, contrary to physical phenomena, understanding computational processes
involved in the execution of programs requires logical analysis of their behavior.
This concerns proving theorems by using the mathematical definition of the pro-
gramming language of the programs and axioms about its data structures. Such an
analysis may lead to the discovery of invariant relationships (invariants) linking
program variables that remain true over program execution. Therefore, each program
is a separate world governed by its own laws, which are indirectly defined by its
programmer. However, they are different from physical laws, as they are immutable
and can be uncovered through logical analysis and proof.

5.1.2 Physical Phenomena as Computational Processes

There are two different approaches to correlating physical phenomena and compu-
tational processes.
The first approach, followed by digital physics, considers that the universe can be
modeled as a computer endlessly executing a program [1]. The states of this program
are defined by the values of the physical quantities observed. The executed program
runs through states that precisely characterize the changes to physical phenomena.
Without going into technical details, there are various ways of digital representation,
one of which—and which I find interesting—uses cellular automata: see Stephen
Wolfram’s A New Kind of Science [2]. A cellular automaton is a potentially infinite
5.2 Comparing Physical Phenomena and Computational Processes 55

set of simple machines (“cells”) arranged in space so that each machine can locally
interact with neighboring machines. While the functioning of each cell can be very
simple, a cellular automaton presents strikingly complex emergent behaviors—see
The Game of Life by John Horton Conway [3].
The second approach, rather than considering the universe to be a computer,
interprets physical phenomena as computational processes that change the values of
physical variables according to physical laws.
This leads to the concept of a natural computer in the following way. Let us
assume that I know the law governing a physical phenomenon—for example, I know
that if I throw a stone from a certain height, it will follow a parabolic trajectory. We
can consider that in this case, the stone is a computer that solves the equation of
parabola in real time.
This very simple idea that every physical phenomenon also involves a computa-
tional process proves to be very fertile. It was initially applied to the so-called analog
computers, which were used to solve linear systems of differential equations.
Analog computers are simple electrical circuits whose behavior is described by
systems of linear differential equations. We can therefore use them to study any
problem that involves solving such equations. All that the “programmer” needs to do
is to build an electrical circuit where each variable of the problem is represented by
voltage at a point in the circuit. Thus, the solution to the problem is the observed
behavior of the circuit: for each variable of the problem; it is given by the function
describing the change in voltage at the corresponding point.
At present, we have other interesting examples of natural computation. These
include neural networks, which mimic the processes of the neural circuits of the
brain, and are used effectively for machine learning; these were previously discussed
under Sect. 3.3.2. There are also quantum computers and bio-computers that lever-
age properties of quantum phenomena and of proteins, respectively, to implement
basic sets of operations.
Natural computers do not execute programs written by programmers. They
effectively compute families of functions whose parameters we can define. They
“mimic” physical systems and are best suited to model physical processes, including
learning, as proven by the achievements of artificial intelligence.

5.2 Comparing Physical Phenomena and Computational

Processes
5.2.1 Computation and the Concept of Continuum

Physical reality cannot be understood without the concept of space-time. Time is the
common parameter of the observed quantities, which change in a “synchronous”
way as its values increase. Each point in space-time defines a “now” and evolution
from each point takes place in parallel as time progresses.
56 5 Physical Phenomena and Computational Processes

Regardless of the nature of the phenomena of the physical world—continuous or

discrete—the assumption that space-time is continuous has proven extremely useful
in classical physics. Without differential equations and integration, it is impossible to
understand most phenomena. Continuous models are best suited for mathematical
analysis and can describe phenomena that discrete models cannot capture.
Perhaps we would have never discovered the idea of continuity if we saw the
phenomena on a different scale, for example, one of millimeters. There are no
straight lines in nature. When a drop of water slides down a glass, it will follow a
notionally straight trajectory that will be distorted by the irregularities of the
surface—after all, there is no such thing as a perfectly smooth surface. The idea of
continuity raises logical problems that have been discussed since antiquity. The
length of a beach is different if measured by a man, a crab, or a microbe! It can be as
large as we want it to be, if it is a continuous quantity. Of course, according to
quantum mechanics, the length is discontinuous—as the minimum measurable
length is defined by the Planck length. However, this does not detract from the
usefulness of continuous models.
An ordered set is continuous if, between any two elements a and b such that
a < b, there is another, different element c such that a < c < b. Such sets are rational
and real numbers. Only continuous mathematics allow for faithful modeling of
physical phenomena. All classical physics and technical culture in general are
based on continuous mathematical models. If we confine ourselves to discrete
mathematics, then we cannot study dynamic phenomena that are usually described
by differential equations.
The idea of continuity in nature was criticized by Democritus and Zeno with his
famous paradoxes. The fable of the race between the swift-footed Achilles and the
slow-moving turtle is well known. Every step Achilles takes is half the distance that
separates him from the finish line, while the turtle is moving at a steady speed. What
Zeno observed is that the turtle will always reach the finish line first—assuming, of
course, that space is continuous. The proof is based on the observation that Achilles
will always be at a certain distance, however infinitesimal, from the finish line due to
the continuity of space.
Continuity is a very useful abstraction. Even if nature contains no straight lines or
perfect circles, we use these concepts to understand space and its properties in a
formal way. An abstract geometric shape is defined using very little information—a
line is defined by two points, and a circle is defined by its center and radius.
Therefore, we believe that we are driving on smooth roads with no discontinuities,
and that the risk of any potholes is marked with appropriate signs. What makes
reality predictable is the smoothness of continuity.
The fact that computers deal only with discrete domains limits their ability to
simulate converging sequences of events described only by continuous mathematics.
Let me explain this by giving an example.
I want to study the movement of a sphere that I allow to fall to a floor from a
certain height. As is indeed the case, every time the ball hits the floor, it loses part of
its energy and, as a result, it eventually stops. Describing this phenomenon using
high-school-level physics, in order to calculate when the sphere will stop, we need to
5.2 Comparing Physical Phenomena and Computational Processes 57

find the limit of the converging sequence of ground collision times. The difference
between successive collision times can become smaller than any finite quantity. No
computer can compute that limit.
If we ask the computer to solve the equations that govern the phenomenon, it will
stop its computation at a point that is, perhaps, very close to the limit. The obvious
reason for this is that it cannot handle infinitesimal quantities that are involved in the
computation of the limit and that are smaller than any measurable quantity but
greater than zero. The computer has a finite memory. An infinitesimal quantity
whose digital representation exceeds its memory capacity can always be found.
A different approach to computing such a limit would be to program the computer
to prove the convergence of the sequence. However, in order to do so, we need to
apply mathematical induction by using rules of logic and algebraic properties. This is
also infeasible because computers cannot systematically discover inductive assump-
tions because of Gödel’s theorems (see Sect. 4.1.1.2, Formalized Languages—
Theories).
In summary, I would say that computers, which are based on discrete models that
do not “have the power” of continuous models, can only approach the accuracy of
the continuous—and this at a significant computational cost. The obvious question
here, of course, is how programmers of simulators of physical phenomena can
realistically represent their dynamics, for example, in video games or even in real-
time control and virtual reality systems.
The answer is that there are techniques devised by people who understand the
dynamics of the systems and program them appropriately so as to solve them.
However, these techniques are specific to each dynamic system. They cannot be
generalized and automated.

5.2.2 Conflicts and Resources in Systems

Discreteness of computing begets phenomena, such as the conflict between actions

to obtain resources. Such phenomena are of particular interest because they also
emerge in economic and social systems, as well as in mental processes, as we will
explain in Chap. 7.
We consider that a system is described by: (1) a set of variables, which represent
its states; and (2) a set of actions:
1. The possible valuations of the variables define the set of the states of the system.
In the case of a mechanical system, the variables are usually continuous. For
example, the state of a moving body is determined by its position, speed, and
acceleration. In the case of a program, the variables are discrete and represent the
program data. A state is defined by a snapshot of the values they take during a step
of execution.
2. An action is a rule governing the change of states. It is possible from a given state
only when the values of the variables characterizing the state satisfy certain
58 5 Physical Phenomena and Computational Processes

preconditions. For example, a thermostat acts by ordering heating to start when

the room temperature (the variable) is, for example, lower than 18 degrees. The
precondition for an elevator door to open is for the elevator to have stopped at a
floor.
The execution of an action, when the relevant preconditions are met, results in
a transition from the present state to a new one, which implies a change in the
values of the system variables. Therefore, in the case of the action of purchasing a
product worth 100 euros using my credit card, the precondition is that the
crediting of an amount of that value is allowed. When the transaction takes
place, my bank account will enter a new state (reduced by 100 euros).
The concept of a resource is common and very important for all kinds of systems.
The existence of sufficient resources is a precondition for actions to take place and
states to change. For an action, a resource is a state variable for which some
minimum value is necessary for its execution, i.e., resources are involved in action
preconditions.
A resource may be quantitative and may represent physical or economic quanti-
ties, such as energy, power, memory, time, physical goods, or money. Therefore,
time and memory are the resources required to execute a program, money to buy a
product and force/energy to accelerate a body.
Furthermore, a resource can be qualitative, i.e., its availability is a precondition
that may or may not be fulfilled. For example, knowledge is a qualitative resource
that enables actions to be carried out: I either know or do not know how to swim,
how to converse in English, how to cook stew, etc. With regard to qualitative
resources, we assume that the corresponding variable has a value of 1 or 0, depending
on whether it is available or not.
Each action needs a minimum amount of resources to be executed, which, in the
case of qualitative resources, is conventionally considered to be 1. When the action is
executed, it consumes the necessary resources and may release other resources. For
example, when a program is being executed, it will reserve the amount of memory
and the processor time necessary for its execution. When the program ends, it
releases the memory used. It should be noted that there are resources that are
consumed when used, such as program execution time.
In other cases, the resource quantity remains unchanged: resources are committed
when used and released after use. These include immaterial resources, such as
knowledge, as well as various body organs whose use is necessary for an action,
such as movement or problem-solving. For example, when I walk, I use my legs,
which are non-consumable resources for the action of locomotion. The chemical
reaction Η2 + Ο ! Η2Ο requires two hydrogen atoms and one oxygen atom to
produce one molecule of water in the presence of a catalyst, which is a
non-consumable resource.
The existence and management of resources is important for the functioning of
systems, as they determine the feasibility of actions and their overall behavior.
However, there is a key difference between physical continuous systems and
computational discrete systems.
5.2 Comparing Physical Phenomena and Computational Processes 59

In physical systems with continuous actions, if two actions taking place at the
same time need common resources in order to be realized, the allocation of resources
per action is carried out in a continuous manner, so as to satisfy the basic laws of
physics.
Conversely, in discrete systems, such as computers and human transaction
systems, it is possible to have a conflict between two actions, which can be described
as follows:
Let us suppose two actions a and b are possible from the state of a discrete system,
and each action needs quantities of a consumable resource, for example, memory or
money, which are ra and rb, respectively, in order to be carried out. If the available
quantity of resources is less than the sum of ra + rb, the execution of an action will
reduce the available resource to such a level that the execution of the other action is
no longer feasible.
In other words, a conflict between two actions a and b can occur when the two
actions need a common resource in order to be executed and the execution of one
disables the execution of the other.
There are countless examples of conflicts not only in computers, but also in
everyday life. When I have a certain amount of money available which is not enough
to fully carry out two actions, then I have to resolve the conflict by making a choice.
Similarly, if the execution of two programs a and b by a computer requires
computation times ta and tb, respectively, and the computer is available for a time
that is less than the sum of ta + tb, then executing one program does not allow the
other to be completed.
In summary, a system has a set of states and actions. States define the available
resources (which are consumable or non-consumable). The execution of each action
from a state depends on preconditions for the resources required. When an action is
carried out, the system enters a new state where the required resources are consumed,
and new resources may be released.
Competition between actions when resources are limited follows a “dog-eat-dog”
logic—the action chosen to be executed takes up the shared resources and precludes
other actions. When two or more actions conflict, information is needed to choose
those that make the best use of resources and avoid deadlocks.
Deadlocks are a very characteristic phenomenon of real-life discrete systems. A
system is in deadlock if it has exhausted all available resources and no action is
longer possible. How can this happen? Deadlocks are a kind of “bankruptcy” of
systems.
Let us assume a computer has 100 memory units available for program execution,
and that this quantity of memory is sufficient to execute any program. In practice, the
computer “runs” several programs concurrently, for example, for graphics manage-
ment, connection to the Internet, various applications, etc. When a program is being
executed, it “asks” the memory management system for an initial quantity of
memory, followed by successive memory requests in proportion to its needs, until
the program ends. The program then releases all the memory it was allocated to be
executed.
60 5 Physical Phenomena and Computational Processes

It is now easy to imagine how a deadlock can arise: when the computer has a total
of 100 memory units and a certain quantity has already been allocated—let us say a
total of 90 units—to the programs already running, and each one needs more than
10 units in order to be able to terminate. The only solution in this case is to stop at
least one program in execution to free up the memory required for the others to
continue being executed. It is clear that deadlocks are dangerous for computers
running critical applications because they may lead to the abnormal termination of
programs.
However, deadlocks arise not only in computational processes, but also in any
system that manages energy, monetary, and even immaterial resources, such as
rights. The insolvency of a company is precisely a form of deadlock. Another
example of a deadlock in the context of a political crisis is a situation where the
current rules do not lead out of this situation—all possible actions are prohibited.
Deadlocks are undesirable situations that block system evolution. This is why
resources must be managed with great care. Computers use resource management
systems that make decisions based on predetermined rules. If there is a conflict
between two actions a and b, the rules generally aim at allocating resources
according to a certain criterion. A “fairness” criterion would require alternation in
the resolution of conflicts between a and b. Other criteria are based on the use of
priorities according to the importance of the actions. Thus, the actions of an alarm
system have a higher priority than other actions taking place under normal circum-
stances. This is also the case with a road junction, which is the resource shared by
vehicles, where traffic on a central road can take priority over a secondary road
intersecting it.
Of course, we can build conflict-free systems by sufficiently increasing the
resources available. We then have a state of abundance where there is no conflict
between the actions sharing these resources. Such systems do not make efficient use
of their resources and are therefore uneconomical: they require the availability of as
many resources as the maximum demand from all their actions.
There is another simple category of systems where, by design, there can be no
conflict: systems whose state unequivocally determines their evolution. These sys-
tems are deterministic, such as the electro-mechanical systems of classical physics.
Given the initial state, the subsequent states are generally defined by the time elapsed
since the system began. Programs with a single “thread of execution” are also
deterministic, i.e., at each step, only a single command can be executed.
In summary, I would say that resource management and deadlock avoidance are
very important problems for discrete systems and computers in particular. This is a
fundamental difference between continuous physical phenomena and discrete pro-
cesses that describe computational as well as economic and social phenomena. In the
following chapters, we will demonstrate that conflict is a key concept for under-
standing consciousness and mental processes.
5.2 Comparing Physical Phenomena and Computational Processes 61

5.2.3 Time and Synchronization

Physical phenomena are described by relationships involving variables that are

functions of time. Much has been written and said about time. Let us confine
ourselves to saying that time is but a common parameter of physical quantities
whose states change as the parameter increases, i.e., time passes. An important
concept here is synchronization: for every single infinitesimal increase in the param-
eter of time, all physical quantities change continuously and simultaneously. It is just
like a movie. The position, speed, and acceleration of a moving body change as a
function of time. When we have many moving bodies in the same frame of reference,
by knowing their initial kinetic state, their positions are determined only by the time
that has elapsed since they began moving. This wondrous property of physical
phenomena at the scale we observe them makes understanding them extremely
simple. As time progresses, everything takes place in a predetermined and parallel
manner within space. This property of determinism, which is linked to the absolute
synchronization of phenomena and which we now consider so “natural,” is not so
easy to capture as a computational process.
The German mathematician and philosopher Gottfried Wilhelm von Leibniz, a
contemporary of Newton and co-inventor of differential calculus, wondered how
two pendulums with exactly the same characteristics could move in an absolutely
synchronous way, without any interaction between their movements. This phenom-
enon is of course, explained in physics by the fact that the pendulums are suspended
in the same gravitational field characterized by g, the gravitational acceleration
constant. Whatever the explanation of the phenomenon according to physics, Leib-
niz was right as a profound thinker and philosopher: what nature achieves without
apparent cost has enormous computational complexity because there is no built-in
notion of space and time in algorithms.
The theory of computation is based solely on arithmetic and is independent of the
physical properties of space-time, which can only be modeled in an approximate
way. If we had programs in the place of the pendulums, their exact synchronization
would entail considerable computational overhead. Moreover, since computation is
discrete, the synchronization would not be perfect as it is in nature.
There are significant differences between physical time and concepts of time we
can associate with computational processes. Of course, in both cases, time helps us
comprehend change in the phenomena.
We have said that physical time can be understood as a monotonously increasing
parameter that “marks” all events in the physical universe. Each event may be dated
depending on the observer. Physical time is sometimes called exogenous time
because it is not possible to stop its passage since change is inherent to physical
reality. It could be thought of as a wheel, which as it turns, determines all changes in
the phenomena. It allows us to correlate the rate of change of phenomena occurring
in parallel in the universe.
62 5 Physical Phenomena and Computational Processes

We have also said that a program computes a function as a logical sequence of

commands and that it is independent of physical time. The execution time of a
program depends on the speed of the computer executing it.
However, when we use computers to simulate physical phenomena, physical time
is represented by a variable that increases as the computer executes the simulation
process. This variable changes according to the simulation speed and can be
characterized as endogenous time because its speed of change bears no relation to
the actual time of the computer. Its value represents the time of the phenomenon
being simulated. Thus, simulating a second in the evolution of a complex phenom-
enon can last days, depending on the speed of the computers executing the simula-
tion program.
Today we use powerful computers to simulate—for the purpose of experimen-
tally studying—the behavior of complex electro-mechanical systems, such as cars or
aircraft when they are being designed. In other words, we use virtual models instead
of building physical models, even on a small scale. These design techniques afford
engineers the advantage of efficiency by both reducing development costs and
allowing experimentation, simply by changing the model parameter combinations.
This is what we call model-based design.
A computer simulation of a complex system with many variables that are
functions of time involves the execution of a circular process that computes the
evolution of each variable by steps that correspond to the change for a short period of
time, a time step Δ. By knowing the values of the system variables (the state of the
system), the computer calculates the new values after time Δ has elapsed. However,
there are phenomena that change very fast and the time step Δ must be variable to
take into account bursts of events. Adjusting the values of Δ to the dynamics of the
phenomenon being simulated is a challenging problem.
The simulation becomes even more difficult when the model is so complex that it
cannot be resolved by a single computer. In such cases, we have to use arrays of
cooperating computers, which requires the synchronization of the simulation pro-
grams being run by all the computers. In other words, the endogenous computation
times of the simulation programs must coincide as much as possible, and this can
entail a computational overhead so large that it can neutralize the advantage of
parallel computing.
Note that the physicist Richard Feynman, who is considered the father of
quantum computing, discussed this problem in 1982. In his article titled “Simulation
Physics with Computers” [4], Feynman says: “The rule of simulation that I would
like to have is that the number of computer elements required to simulate a large
physical system is only to be proportional to the space-time volume of the physical
system. I do not want to have an explosion.”
Without going into technical details, I would say that the simulation of physical
phenomena by computers involves high complexity due to the discrete nature of
their computation and the cost of synchronizing their simulation processes. This is
something that physical time achieves in a completely “magical” way, with no cost
whatsoever.
5.2 Comparing Physical Phenomena and Computational Processes 63

5.2.4 Why Is the Physical World Comprehensible?

A general observation we can make when comparing the physical world to the world
of computers at the level where we observe them is that the physical world has
admirable uniformity and normality in its structure and behavior. The same basic,
simple laws govern physical phenomena everywhere in space-time. In contrast, each
program can be considered to be a universe with its own particular laws. These laws
are generally complex and are very difficult to guess experimentally. As we have
said, there are techniques that allow us to discover them: given a program, one can
theoretically compute invariant relationships between its variables.
If the physical world were built like the world of programs, then there would be
no science whatsoever and scientific truths would go out the window.
We have explained that physical laws resulted from a process of generalization,
which is logically arbitrary. Humans discovered them experimentally. For example,
we found that the ratio between the current and the voltage of a resistor is constant
and that the air resistance of a moving body is proportional to the square of its speed.
However, as I have explained, this process of observation/generalization in order to
discover the laws governing physical phenomena does not work in computing
systems. On the one hand, I have a “uniform” behavior captured by simple laws,
while on the other hand, I do not have any such behavior.
Why can we not have discrete computer systems that behave “normally,” such
that, after a relatively large number of experimental observations, we can risk
making a generalization with high likelihood as we would do for simple physical
systems?
Let us make another comparison. The physical world is a combination of
particles, while the computer world is a combination of computational elements.
How can it be that, at a high layer of the hierarchy in the physical world, simple and
predictable laws emerge, while the same is not true in the computer world? Regard-
ing this wondrous property of the physical world, Einstein said that “the most
incomprehensible thing about the universe is that it is comprehensible.” The basic
laws of physics are astoundingly simple and normal cause-and-effect relationships
can be characterized by simple linear laws. This allows us to approximate physical
phenomena fairly faithfully using systems of linear differential equations. Of course,
there are also more complex and chaotic phenomena. Small changes to certain
parameters can change the entire dynamics of the system—the “butterfly effect”
comes to mind. However, all technical civilization is based on the fact that the
electro-mechanical artifacts we build adhere to extremely simple laws that allow for
predictability.
Computing systems, even in their most elementary form, are difficult to under-
stand. The vast majority of such systems are nonlinear, such as the simplest system,
which is a memory component. Furthermore, computing systems are, by their very
nature, chaotic due to the discrete nature of information. Can we have programs that
behave robustly? Can we find a programming language where small changes in the
program’s code would result in a small difference in their behavior, as is the case
64 5 Physical Phenomena and Computational Processes

with small changes in the parameters of an electro-mechanical system, which

generally do not bring about dramatic changes in its behavior?
Nature does in fact possess a robustness that is alien to computers. If, for example,
a civil engineer designs a building whose structural stability has been calculated for
steel of a certain strength, it is certain that the building will be structurally sound if
higher strength steel is used. This does not apply to computer systems, which are not
resilient in this way. What we call “anomalies” appear: the improvement of certain
parameters of a system’s components does not bring about the reasonably expected
improvement in the overall behavior of the system.
A well-known example concerns time anomalies, where the following “paradox”
occurs: When one executes a program on a computer, one reasonably expects that
the faster the computer, the shorter the execution time. There are cases where this is
not true. Performance not only does not improve but may actually deteriorate.
These anomalies are due to the non-determinism of computational processes and
the way in which conflicts are resolved, as discussed above. When we try to improve
system performance by increasing the resources available, the degree of freedom in
resource management increases and new choices emerge, which could lead to
devastating results.
Similar anomalies occur when we increase the available memory. It is known that
lack of available memory during the execution of a program can lead to deadlock.
However, in certain cases, increasing the quantity of memory in a system function-
ing normally can also lead to deadlock.
This discussion indicates that there are significant differences between physical
and computing systems. These differences are due to the discrete nature of the latter,
whose behavior and properties present interesting similarities to economic and social
phenomena.

References

1. https://en.wikipedia.org/wiki/Digital_physics
2. https://writings.stephenwolfram.com/2017/05/a-new-kind-of-science-a-15-year-view/
3. https://en.wikipedia.org/wiki/Conway%27s_Game_of_Life
4. http://research.physics.illinois.edu/DeMarco/images/feynman.pdf
Chapter 6
Human vs. Artificial Intelligence

The relationship between human intelligence and artificial intelligence is a very

topical and much debated issue due to recent advances in machine learning and its
spectacular applications in the development of smart services and systems.

6.1 Human Intelligence Characteristics

6.1.1 Slow and Fast Thinking

It is a well-known fact that our mind combines two ways of thinking [1]: (1) slow
and conscious thinking, which is procedural and applies the rules of reasoning, and
(2) fast and automatic (non-conscious) thinking.
We use slow thinking to consciously solve a problem, for example, to analyze a
situation, to plan an action or to build an artifact. Fast thinking is automatic and
allows us to solve problems whose complexity is extremely high, such as when we
talk, walk, play the piano, etc. When we walk, our “fast computer” solves a problem
whose algorithmic real-time solution would be extremely complex. Understanding
and analyzing fast thinking is extremely challenging. If a pianist playing a difficult
composition uses conscious thinking in an effort to analyze exactly what she is doing
at a given time, she may become confused and lose her pace.
Automatic thinking is the most important factor in human intelligence. If we
compare the information processing capacities of the two systems, we see that a large
part of its bandwidth is due to fast thinking. Of course, this comparison is quanti-
tative and does not reflect the role and importance of conscious thinking for humans.
There is also remarkable cooperation and complementarity between the two
modes of thinking. Following birth, the conscious and slow system invents processes
that are automated by the fast system. When an infant babbles and is consciously
saying “ma-ma” the fast “computer” learns progressively until it takes control of the
function of language. The same happens when we learn how to ride a bicycle. First,

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 65
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_6
66 6 Human vs. Artificial Intelligence

we consciously try, following a “trial and error” process, to balance on the two
wheels. Ultimately, the fast “computer” learns how to balance automatically, without
us even understanding the laws of mechanics.
There is a stunning correspondence between the two modes of thinking and the
two basic models of computation. Conscious thinking is based on mental models that
we understand and is what we use to program computers. Automatic thinking is the
result of learning, just like the computation of neural networks.
If we want to build a bipedal robot or a robot riding a bicycle, we can take two
different approaches: we will either program the computers controlling the move-
ment of the robot, or we will use a suitably trained neural network. In the first case,
we must use the theories of mechanics and write the real-time control programs. In
the second case, we must train the neural network to learn how to balance, exactly as
we learn without using theoretical knowledge.
Mathematics and logic, being creations of conscious thinking, reflect how it
works. The algorithms and programs that computers execute are but formalized
processes of this mode of thinking.
Incidentally, I would like to stress that what we call “causality” is not necessarily
a property of physical phenomena. It can simply reflect the way we understand
phenomena because conscious thinking is serial, and the rules of reasoning formalize
preservation of truth as a cause-and-effect relationship. Furthermore, computational
processes involve causal chains of operations where, when an operation is com-
pleted, its results trigger subsequent operations.
Of course, in mathematics we can formulate relationships, that are not causal, and
therefore not executable as a single computational process, but that characterize sets
of computational processes that are capable of multiple dynamic correlations.
Natural computers that rely on physical processes directly, and neural networks in
particular, are best suited to model fast thinking. They are inherently parallel and can
perform computation whose analysis using logical processes is practically impossi-
ble. Unfortunately, fast thinking is non-conscious and, therefore, understanding and
analyzing its basic laws and mechanisms would be a hopeless endeavor.
An interesting question arises: Which functions can natural computers effectively
compute? Is it possible to have a new theory of computation that uses the compu-
tational capabilities of physical systems, for example, analog, neural, quantum?

6.1.2 Common Sense Intelligence

I have always said that any “intelligence” that conventional computers possess
reflects the intelligence of their programmers. They simply execute commands
describing the mechanical processing of symbols.
Of course, things have changed with the advent of neural networks, which follow
a radically different computational model that is not programmed but simply learns
from very large datasets. However, we certainly have “smart systems” that specialize
in solving one type of problem: they can play chess, classify images, take part in
6.1 Human Intelligence Characteristics 67

television game shows, etc.—but a system that is a chess grandmaster will fail badly
at driving a car.
One step toward approximating human intelligence would be to have computers
that exhibit what we call strong AI. In other words, they could combine solutions to
specific problems and skills, exactly as people do, by responding to stimuli from
their environment.
What characterizes human intelligence is the combination of perception/interpre-
tation of sensory information, their logical processing and decision-making that
could potentially lead to action. Such behavior has nothing to do with the develop-
ment of strategies for playing chess or go. In games, the rules are predetermined and
do not change. Humans are able to adapt to environments where rules and goals
change dynamically depending on circumstances.
As I will explain in what follows, we can understand consciousness as our mind’s
ability to “see” ourselves acting in a semantic model of the external and internal
world. This model is, at first, built automatically from our infancy. It is then
consciously enriched through learning. It can be seen as a dynamic system with
states that reflect, on the one hand, our perception of the external environment and,
on the other hand, our awareness of our internal state. Without such a semantic
model, it would not be possible for us to understand languages and therefore
communicate with each other. At the same time, the use of language plays an
important role in correlating the semantic models each of us develops individually.
Much of human intelligence is due to what we call common sense. Our minds can
use their semantic model to assess a situation and the consequences of what is taking
place in our environment. The model is created by the accumulated experience and is
almost automatically enriched every day. Thus, the “parent–child” relationship is
enriched by a set of other relationships that common sense uses, such as age,
hierarchy, and support, which are very difficult to list and formalize.
For computers to exhibit this type of behavior, we must endow them with a
corresponding semantic model. In theory, we could build such a model if we could
analyze and formalize natural languages; if we could model them with hierarchically
structured semantic networks of relationships between concepts, equipped with rules
for representing and updating knowledge. For example, when defining the “parent–
child” relationship, we have to imagine and formulate all the relevant relationships
and rules it entails. Unfortunately, despite research efforts for more than 50 years,
very little progress has been made in this direction.
For example, humans interpret the sequence of images shown in Fig. 6.1 almost
instantaneously as an aircraft accident. This is because the process of understanding
can link the contexts of the perceived sequence of images using common sense
knowledge of the semantic model. In contrast, a computer can analyze the sensory
information for each image but lacks the knowledge that would allow it to reach the
same conclusion. It can identify and perhaps correlate the objects appearing in a
single image, but it is unable to analyze and understand the dynamic relationship
between images.
68 6 Human vs. Artificial Intelligence

Fig. 6.1 Humans are endowed with common sense reasoning

6.1.3 Cognitive Complexity—The Boundaries

of Understanding

Understanding means to be able to link an observation to a relationship one has

mentally mastered. We can realize the proportionality between force and accelera-
tion or the exponential growth of an organism over time because we have already
mastered the relevant concepts and their properties at school or through experience.
When we study phenomena affected by a large number of parameters, we are
forced to make certain simplifications based on assumptions that some parameters
are primary and others are secondary. Such simplifications are not always legitimate.
For example, to consider phenomena without friction in many cases simplifies
theory. There are cases, however, where friction is essential to the dynamics of a
phenomenon and cannot be ignored.
There are phenomena that are inherently complex and cannot be simplified, such
as meteorological, economic, and social phenomena. The number of parameters that
need to be understood is quite large. Therefore, we cannot study these phenomena in
theory, not because they cannot be causally understood, but because the discovery of
relations between observations involves an inherent complexity that the human mind
cannot grasp.
It is well known that the human mind is constrained by cognitive complexity,
which we define as the time needed to grasp a relationship. It has been experimen-
tally determined that the limit to the number of parameters that our mind can
correlate is approximately five (one relationship with four parameters). This hin-
drance to conceiving relationships with many operands is a significant constraint that
affects our ability to understand complex phenomena, but also limits the complexity
6.2 Weak Artificial Intelligence 69

of the theories and artifacts that we can build. I noted above that Einstein was right to
observe that, for our good fortune, the basic laws of physics are astoundingly simple.
The existing scientific theories involve a relatively small number of independent
quantities and concepts to formulate these laws. I can also say from personal
experience that complex theories are difficult to understand and therefore difficult
to use, while checking their validity often proves problematic.
Therefore, developing theories that would allow a holistic approach to the study
of complex phenomena comes up against the constraints of the human mind. It is
here that I believe cooperation between humans and computers could help us
overcome these restrictions, to a certain extent, as I will explain in what follows,
after discussing the validity of machine-generated knowledge.

6.2 Weak Artificial Intelligence

Living organisms and humans, in particular, can be considered to be computing

machines that bear certain similarities to computers: they use memory and lan-
guages. The correspondence between hardware/software and brain/mind is also
noteworthy.
However, there are also some important differences. Computational phenomena
in living organisms are “resilient”—they have inherent adaptability mechanisms
and, most importantly, have allowed for the emergence of concepts and language.
Computers exceed human thinking in being able to compute faster and more
accurately. This enables them to compete with humans in solving problems that
require the examination of an unfeasibly large number of solutions or a combination
of a large number of predetermined data. This idea is applied in systems such as
IBM’s Deep Blue and Watson and Google’s AlphaGo. The fact that human intelli-
gence is defeated in such games leads some people to believe that computers are
“smarter” than humans.
It is worth briefly discussing the evolution of artificial intelligence, which was
born in the mid-1950s as a branch of informatics that “studies and designs intelligent
systems.” Since the mid-1960s, its proponents would enthusiastically promise that
they could build machines that would rival human intelligence, as two examples of
declarations from that period show: “machines will be capable, within 20 years, of
doing any work a man can do”; “within a generation . . . the problem of creating
‘artificial intelligence’ will substantially be solved” [2]. Of course, such promises
were not kept for many reasons. One is due to a lack of understanding of the
limitations of the theory of computation. Another, more profound reason is that in
order to build a machine that behaves like a human, we must have understood and
analyzed the mechanisms of human intelligence.
The focus and perimeter of artificial intelligence has shifted over the years, with
corresponding fluctuations in the interest and funding of research. There are two
characteristic periods known as “AI winters”: one in the early 1970s and one in the
late 1980s.
70 6 Human vs. Artificial Intelligence

Since the beginning of the twenty-first century, there has been a rekindling of
interest, particularly thanks to achievements in machine learning and data analytics.
This is accompanied by a “frenzy of optimism” regarding the importance and impact
of artificial intelligence, particularly on the development of autonomous systems and
services.
Many still consider that intelligence is limited to solving complex but well-
defined problems by focusing on decision-making. They believe that, thanks to the
use of machine learning methods, they will be able to rise to the challenge.
I believe that what characterizes human intelligence is autonomous behavior and
adaptation to changes in the internal and external environment. It is the incomparable
ability of the human mind to create new knowledge, to understand situations one has
never encountered before, and to set new goals. The gap between human and
artificial intelligence will be bridged when we have computerized systems that
autonomously perform a large number of services while adapting to changing
situations.

6.2.1 The Turing Test

In order to judge whether a computer A is just as intelligent as a person B, Alan

Turing proposed a test. The test consists of having their behavior compared by an
interrogator C, who sends questions to A and B and who, in turn, provide a
corresponding answer for each question (Fig. 6.2). If the interrogator cannot tell
from the answers, which is the computer and which the person, then it is concluded
that they are equally intelligent.
The relevance of the Turing test was challenged by philosopher John Rogers
Searle, who proposed the Chinese Room Argument. This is a thought experiment,
which shows that programming a computer may make it appear to understand
language by manipulating strings of symbols without understanding their meaning.
The experiment, the layout of which is shown in Fig. 6.2 to the right, assumes that
B—whether man or computer, it does not matter—is in a room receiving a number of
questions written in Chinese by A. The experiment intends to prove that there is a

...................... ......................

Fig. 6.2 The Turing test (left) and the Chinese room argument (right)
6.2 Weak Artificial Intelligence 71

way for B to give C the right answer to every question asked by A without speaking a
word of Chinese. Therefore, the Turing test, as well as all other tests comparing
behaviors, is not appropriate.
Let us assume that B has access to a huge database containing all possible
questions that can be worded in Chinese and their respective answers in the same
language. This is theoretically feasible, as there is but a finite number of questions in
Chinese.
Thus, when B receives a question, B will search the database, find the
corresponding answer and give it to C. It is obvious that this process can be
automated by ordering the questions lexicographically. This thought experiment
shows that, if we take observed behavior as a benchmark, we may be led to
erroneous conclusions.
Another argument concerning the Turing test is that a set of specific questions
may prove to be the undoing of a human participant. For example, if the interrogator
asks “What is the hundredth digit of the decimal expansion of π?”, the computer can
respond immediately, while the human will be unable to answer relying solely on his
or her mental faculties. Thus, the computer is superior to humans when facing this
type of question!
Note that we understand the world by observing and studying behaviors. There-
fore, the only methodologically legitimate criterion is comparing behaviors, and we
have seen that this is not fit for purpose.
However, if we use ontological criteria, a distinction is feasible. Without knowing
exactly what the mind is, we can say with certainty that the computer is not a mind
because the mind can build a computer, whereas the opposite is impossible given the
present circumstances.
A closing methodological remark that will be further discussed in Sect. 6.3
concerns autonomous systems. The criterion for intelligence cannot be reduced to
a Q&A game; it should be about building systems capable of replacing human agents
performing tasks in complex organizations. This is a much more ambitious goal than
the one set by the Turing test.

6.2.2 Machine Learning and Scientific Knowledge

Recent advances in machine learning and data analytics have shown that it is
possible to generate knowledge and forecasts with some success. Machine learning
and statistical analysis techniques allow for the recognition of complex relationships
in data. These correlations may indicate cause-and-effect relationships and allow for
predictability. However, they may also be random, which is difficult to establish, but
I will not discuss the matter further [3].
How valid is the knowledge generated by neural networks and how much can we
trust it? A comparison of the way it is generated, and the way scientific knowledge is
developed allows us to highlight key differences.
72 6 Human vs. Artificial Intelligence

Fig. 6.3 Processes for producing scientific and machine learning knowledge

As we have said, the scientific method enables the development of knowledge

that explains our observations regarding the world. Scientific discovery is the result
of the learning process of an observer, who formulates a hypothesis by creating a
model that allows for the explainability and predictability of a phenomenon.
On the other hand, neural networks can generate knowledge as the result of a long
“learning” process involving a large dataset. By learning the data representing a
cause-and-effect relationship, the network can estimate the most likely result for a
given cause by applying a kind of extrapolation method. The success rate depends on
the degree of “training” of the network, but we cannot be certain that the response is
correct. Furthermore, our inability to use models in order to understand exactly how
the network works does not allow us to estimate the possibility of an incorrect
response.
Figure 6.3 shows differences between the scientific approach to studying a
physical phenomenon (acceleration a of a mass m from a force F) and the process
whereby a neural network learns a cognitive process (how to recognize images of
cats or dogs). In both cases, there is a common goal: to capture the input-output
function characterizing a cause-and-effect relationship. In the first case, it is accel-
eration a resulting from a given force F, while in the other, it is “cat” or “dog,”
depending on what the input image depicts. In both cases, there is a learning stage at
first.
In the physics experiment, the experimenter, let us say Galileo, learns by relating
cause and effect. By using his capability of mental abstraction and creativity, he will
assume proportionality in this relationship. This hypothesis will ultimately become a
“law” since it is experimentally validated.
6.2 Weak Artificial Intelligence 73

Using a mathematical model allows checking the validity of knowledge (see Sect.
3.4.2). We can use the model to examine the physical phenomenon and see how it
behaves in corner cases and to generally gain insights, thanks to mathematical
analysis.
Similarly, machine learning involves an initial experimental process, where
images are labeled by an experimenter. Then the neural network is “trained” by
adjusting its parameters so that for each image in it provides the correct response rn.
The difference with the scientific approach is that we cannot characterize the input-
output behavior of the neural network through a mathematical model. This would
require, for this particular example, an almost impossible formalization of the
concepts of cat and dog.
However, note that when the network inputs and outputs are physical quantities,
there is no theoretical limitation in this respect. By knowing the structure and
behavior of the components of the neural network, I can theoretically calculate the
mathematical function that characterizes the input-output relationship.
Neural networks are particularly useful where mathematical models cannot help.
This is where mathematics finds itself at a loss, as we do not know how to define
theoretically, what an image means. This is just as difficult as formalizing natural
language.
Nonetheless, the ability of neural networks to deal with non-formalizable knowl-
edge gives rise to an interesting complementarity with conventional computers that
will prove very fruitful in the future.

6.2.3 A New Kind of Knowledge—Prediction Without

Understanding

Can computers help us overcome the limitations that our human nature imposes on
the search for knowledge? The answer is clearly yes. There are numerous examples
where computers contribute to the analysis of complex phenomena, thanks to the
combined use of data analytics and machine learning. It is perhaps, thanks to the use
of computers, that we will be able to overcome the barrier of cognitive complexity,
which was discussed under Sect. 6.1.3. Computers may help us develop and verify
new theories to explain complex phenomena.
Thus, we have a process for generating a new type of scientific knowledge, where
the law is not a clearly formulated mathematical relationship devised by the human
mind, but a possibly complex relationship discovered with the use of computers. The
analysis of this relationship may allow predictability, but it will certainly limit the
deeper understanding of phenomena that would allow their characterization by
explicit mathematical models.
It is, therefore, thanks to computers that we have a new type of knowledge that
allows prediction without the understanding enabled by the power of mathematical
analysis.
74 6 Human vs. Artificial Intelligence

The use of artificial intelligence and supercomputers is paving new roads for the
development of knowledge. This was the argument made years ago during a
discussion with a seismologist acquaintance of mine, who told me that perhaps in
a short while Google would be better at predicting earthquakes than experts. I do not
know whether this will come to pass. However, I think that new techniques for
analyzing large datasets can increase the possibilities of predicting complex phe-
nomena, without necessarily understanding their nature in the absence of explicitly
formulated laws. Taking earthquakes as an example, if we could correlate seismic
activity in one location to seismic activity around the planet, then perhaps we could
make predictions without or with very little theory.
Of course, we must weigh the impact of using such knowledge. The “cloud” is
becoming a type of oracle that people turn to in search of solutions to the complex
problems they face—is this good? bad? dangerous?
The generation and use of knowledge through artificial intelligence techniques
that allow for predictability without understanding, especially if used to make critical
decisions, should give us pause.

6.3 Beyond Weak Artificial Intelligence—Autonomy

6.3.1 Autonomous Systems

I have explained that what distinguishes computers from humans is that current
computers automatically execute certain predetermined functions, whereas humans
can act autonomously; that is, they can act reactively to changes in their environment
and also act proactively, driven by internal goals.
Today, the use of artificial intelligence allows us to take an important step
forward, to progress from automated systems to autonomous systems that are called
upon to replace people in complex functions. The main motivation is to achieve
greater efficiency by combining already automated processes without human inter-
vention. The human factor will only regulate certain goals, delegating their realiza-
tion to autonomous systems. For example, in the case of self-driving cars, we will
simply enter a destination; in the case of smart factories and farms, we will simply
enter production indices.
Achieving autonomous systems and services is a core goal of the Internet of
Things. If this goal is achieved, then we will have much more convincing proof of a
certain kind of computer intelligence that is far more important than beating people
at games.
At this point, the difference between automation and autonomous systems must
be explained. To this end, I will present five different systems in order of difficulty of
design: a thermostat, an automated shuttle, chess-playing robots, football-playing
robots, and self-driving cars (Fig. 6.4).
The common characteristic of these systems is that they use computers to control
their environment, so that their behavior meets certain goals. Computers receive
6.3 Beyond Weak Artificial Intelligence—Autonomy 75

Fig. 6.4 Automated and autonomous systems

information about the state of the environment via sensors, and compute commands,
which they send to actuators that perform the appropriate actions to achieve the
goals.
A thermostat controls the operation of a heater in order to keep the temperature of
a room between a maximum and a minimum value. When the temperature given by
the sensors reaches the minimum value, it commands the heater to begin functioning.
When the temperature reaches the maximum, it orders the heater to stop.
An automated shuttle has a more complex control system. A system that enforces
a series of predetermined stops at stations with corresponding acceleration and
deceleration rates. The system takes into account sensor signals that determine the
position of the shuttle along its path. Designing such a system does not pose any
particular difficulty in its core principles. Particular attention should be given to how
to accelerate and decelerate in order to ensure the safety and comfort of passengers.
A chess-playing robot has a relatively simple environment, whose state is deter-
mined by the position of the pawns on the chessboard. However, its control system is
extremely complex and impossible to design, similar to the two previous systems.
The reason is the astronomically large number of combinations on the chessboard,
and the even larger number of moves for each combination in order to achieve the
game’s objective. Consequently, movements cannot be determined statically (a
priori), but are calculated dynamically. For each state, the robot uses pre-existing
knowledge to calculate tactics, i.e., sequences of actions, which, depending on the
reactions of the opponent, lead to the best result.
A football-playing robot faces an even more complex environment defined by the
players’ positions and speeds. A major difference from the previous example is that
the environment changes dynamically. Therefore, the football-playing robot must be
76 6 Human vs. Artificial Intelligence

able to monitor changes in the environment so as to react successfully in real time.

This means that images from its cameras must be analyzed precisely and in time in
order to depict the state of the pitch as faithfully as possible. Furthermore, the robot
control system dynamically calculates its goals according to its role and position on
the pitch. For example, there are goals that involve defensive behavior and others
that involve offensive behavior. For each specific goal, the robot calculates the
respective tactics over a certain length of time, taking possible reactions by its
opponents into account.
Finally, self-driving cars are certainly the most complex system for many reasons.
First, the physical environment is changing dynamically and is not limited to a pitch
or a chessboard. Moreover, the number of all the vehicles and the position of the
obstacles around it is also changing. The layout of the environment depends on
geography and the infrastructure available (traffic regulation sensors and equipment,
communication networks). Finally, each vehicle has an extremely complex control
system. It uses computers that manage multiple goals to shape a strategy consisting
of selecting the most important goals first for safety and second for comfort. When it
selects a set of compatible goals, it calculates corresponding tactics to achieve them.
The above comparison shows the significant differences between automated and
autonomous systems that are in a way, called upon to demonstrate mental faculties
commensurate to those of humans. The thermostat and train are simply automated
systems because they perform predetermined services with fixed goals in a well-
defined environment.
The other three systems are characterized as autonomous, because they have the
capacity, similar to living organisms, to generate and manage knowledge with a dual
purpose: on the one hand, to “understand” their external environment and, on the
other hand, to adaptively manage multiple goals and plan corresponding actions to
achieve them.

6.3.2 Characteristic Functions and Organization

of Autonomous Systems

Following the above comparison, I propose below an architecture for an autonomous

system that clearly shows how five key functions work together to achieve auton-
omy. A similar model can also be used to reach a theoretical understanding of how
the basic mental functions of consciousness can be combined (see Sect. 7.2.2).
An autonomous system accepts and processes sensory information from the
environment and computes commands whereby actuators carry out actions that
change the state of the environment. For example, in Fig. 6.5, the system is the
autopilot of a self-driving car. The internal environment is the vehicle whose
direction and speed are controlled by the system. In the external environment, we
see three vehicles, one pedestrian and one traffic light. The system processes the
information concerning the environment, both internal and external, and issues
6.3 Beyond Weak Artificial Intelligence—Autonomy 77

Fig. 6.5 Architecture of an autonomous system with its five key functions

commands to carry out actions. These actions must take place within predetermined
time limits defined by the dynamics of the environment in order to achieve the goals
in time.
An autonomous system combines five key functions, two of which aim at
understanding environment situations (perception and reflection) and two at
decision-making (goal management and planning). The fifth function confers the
ability of self-learning.
The system is also equipped with a knowledge repository, where it stores acquired
knowledge that is useful for identifying and managing sensory information, in
particular. Knowledge includes, first and foremost, concepts concerning objects in
the environment and their properties, as well as methods for decision-making. In the
example at hand, the concepts of “car,” “pedestrian,” and “lights” are required to
“understand” the external environment. The repository for each of these concepts
may contain information concerning their characteristic properties so that the system
achieves better predictability, for example, it may know the maximum speed and
acceleration of a specific type of car.
The function of perception receives sensory information from the environment
(image, signals) and analyzes it by distinguishing between concepts and, possibly,
relationships linking them that are stored in the repository. Therefore, in our
example, the sensory information from the external environment contains three
cars, one pedestrian, and one traffic light with corresponding information on their
position and their kinetic state. Regarding the internal environment, the sensory
information concerns the kinetic state of the vehicle, such as speed and acceleration.
78 6 Human vs. Artificial Intelligence

The function of perception is usually accomplished using neural networks, which

are currently the only technology fit for this purpose.
The perceived information is transferred to the reflection function, which is
tasked with building a model of the external and internal environment of the system.
This model has state variables that represent states of the environment such as the
kinetic attributes of the obstacles and the state of the vehicle. The actions of the
model are changes in the states that the controlled vehicle or the obstacles around it
can perform.
The model must be updated in real time so as to reflect the dynamic state of the
environment as closely as possible.
The environment model is used by a decision-making module that combines two
functions.
The first function performs goal management by selecting, from a set of
pre-defined goals, a subset of compatible goals in relation to the current state of
the environment model. Goal management determines what we call the system’s
strategy.
The system’s goals are divided into negative and positive goals.
Negative goals concern the avoidance of undesirable states, such as safety goals
regarding collision avoidance.
Positive goals concern the achievement of desirable conditions, such as optimiz-
ing passenger comfort, fuel consumption, and moving from one location to another.
We can also distinguish among short-term goals, such as safety goals, medium-
term goals, for example, for maneuvering the vehicle to overtake other vehicles or
drive through intersections, and long-term goals, such as completing an itinerary.
Timely and critical goal selection is crucial to system autonomy because it is
highly complex and requires a computation time that must satisfy real-time response
requirements.
Goal management is complemented by the planning function, which determines
the system’s tactics. For each set of selected goals, this function calculates a
sequence of commands to the actuators, which carry out corresponding actions to
realize these commands. Thus, with regard to the collision avoidance goal, it must
control speed and direction by adequately combining braking, acceleration, and
steering wheel angle. For every kind of maneuver, the system must have the
appropriate tactics to achieve its goals.
Finally, the fifth key function is self-learning, which manages and updates the
knowledge on the repository. Knowledge is updated through the creation of new
knowledge concerning: (1) the environment, for example, new concepts based on
accumulated knowledge from the analysis of model data; and (2) new goals for
adapting to changes in the environment or changing parameter values that are
relevant to the choice between goals.
The self-learning function is a key feature of human autonomy. When someone
has never driven on snowy roads, they can adapt their behavior by carefully testing
various tactics and with the goal of minimizing risk. The self-learning potential of
actual systems is limited to parameter optimization but cannot create completely new
concepts or new goals.
6.3 Beyond Weak Artificial Intelligence—Autonomy 79

Note that the functioning of the system is cyclical. The cycle begins with
perception and continues with reflection that updates the environment model. This
is followed by decision-making, possibly with the choice of new goals, and the
implementation of tactics not been completed during the previous cycle. This is
because the cycle duration must be short enough to achieve some short-term goals
(of the order of a tenth of a second for safe driving), while achieving long-term goals
may take millions of cycles (to reach a destination). Obviously, when new goals are
selected in a cycle, they must be compatible with those already selected and not yet
realized.
The above architecture defines autonomy as the capability of a system to achieve
a set of coordinated goals without human intervention, adapting to changes in the
environment. In order to achieve autonomy, it is necessary to combine the five
mutually independent functions: perception, reflection, goal management, planning,
and self-learning.
This characterization allows us to distinguish between automated and autono-
mous systems.
A thermostat is an automated system because it does not need any of the five key
functions.
A shuttle is primarily an automated system although modern versions use the
function of perception to analyze images.
The other systems can be considered as autonomous, but with graduated
difficulty.
For the chess-playing robot, the perception and reflection functions are relatively
simple, as the environment and its possible changes are slow and well defined.
Decisions are also made based on clearly defined rules. The difficulty lies in
computing successful tactics anticipating the opponent’s tactics.
For the football-playing robot, the perception and reflection functions are quite
complex due to the dynamic nature of the sensory information. The strategy also
dynamically changes in cooperation with the other robots on the same team. The
goals may be offensive at times and defensive at others, depending on the robot’s
position in the game. The development of tactics through cooperation between the
players is also dynamic. In order to coordinate the functions, the use of knowledge is
important and includes the rules of the game, as well as knowledge gained dynam-
ically by learning the characteristics of the players, especially those on the
opposing team.
Finally, autonomous cars are among the most difficult systems to achieve due to
the complexity of their environment and the need to manage multiple goals and
adapt in real time.

6.3.3 Should We Trust Autonomous Systems?

The development of reliable autonomous systems is essential in order to realize the

vision of the Internet of Things. It is both a scientific and a technical challenge, with
80 6 Human vs. Artificial Intelligence

tremendous economic and political stakes. That is why all major technology com-
panies are investing in this sector, focusing particularly on self-driving cars, because
of the considerable economic incentives. This battle is being fought by companies
such as Google and its subsidiary Waymo, Apple, Intel and its subsidiary Mobileye,
Uber, the Chinese companies Alibaba, Huawei, Tencent and, of course, all the major
motor manufacturers, with Tesla at the forefront.
The solutions promoted by industry mainly rely on the use of neural networks, as
the algorithmic techniques performed by conventional computers seem to be unable
to solve the problem of perception effectively.
Here I must point out that the use of neural networks in critical autonomous
systems has been the subject of fierce debate, as it raises serious safety problems.
Until recently, the construction of critical systems had to be based on scientific
knowledge and the use of mathematical models. When we build a bridge or design
the autopilot of an aircraft, the use of mathematical models allows us to predict how
these systems behave in various scenarios and to guarantee their safe use with a very
high probability, for example, for a civilian aircraft, the non-hazardous failure rate
per hour of flight will be less than 10-9. I must also stress that all artifacts are
developed on the basis of regulations and standards controlled by independent
certification bodies. When you buy a toaster or car tires, an authority has tested
and guaranteed, based on theoretical and experimental data, that if you use them
correctly, your life will not be at risk.
Unfortunately, machine learning systems are not based on models, but on “accu-
mulated” empirical knowledge. We can, of course, ascertain experimentally whether
they are working properly. However, even if they work perfectly according to a large
body of experimental data, we cannot say with the confidence afforded by the
scientific method that they will continue to work properly (see Sect. 6.2.2). The
certification methods applied by independent organizations require explainability
that cannot be achieved by machine learning systems.
Today, in order to avoid stopping the development of autonomous systems with
neural networks and to maintain their national supremacy, the competent US author-
ities accept their use after “self-certification.” In practice, this means that the
manufacturer—and not an independent authority—guarantees their safety and is
held fully liable in the event of an accident. Thus, if the autopilot system of a car
fails, the manufacturer must pay compensation for any damage and victims.
This policy change, which does not impose objective safety criteria and allows
every manufacturer to act freely, poses serious risks. Human life could simply
become the adaptation parameter of an equation, where economic and technical
criteria are set off. In my view, the use of autonomous systems should be limited
when we cannot guarantee an adequate reliability level.
6.4 Artificial Intelligence—Threats and Challenges 81

6.4 Artificial Intelligence—Threats and Challenges

I will not be discussing the possibilities offered by the use of computers and artificial
intelligence, in particular. They are numerous and relatively well known. The media
often talk about them and the radical changes in the way we live, work, and learn.
The automation of processes and services offers the advantage of efficiency. Without
direct human intervention, we can have “real-time” control, so to speak, of the use of
resources, in sectors such as energy, telecommunications, and transport, in the best
way possible so as to achieve economies of scale and quality of life.
I will discuss the risks, whether hypothetical or real, in greater depth.

6.4.1 Hypothetical Risks

Mythology has recently grown around the issue of computer “ultra-intelligence.”

According to one version, computer intelligence will eventually exceed human
intelligence and perhaps we will end up as something like pets for machines.
These views on the risk posed by the intelligence of machines are espoused by
celebrities such as Stephen Hawking, Bill Gates, and the entrepreneur and inventor
Elon Musk. Some are influenced by the ideas of Ray Kurzweil that the technological
singularity is imminent, which is supposed to happen when the computing power of
the machines outpaces the “computing power” of the human brain [4].
Obviously, all these arguments are flimsy and lack seriousness. Having very
powerful machines is not enough to overcome human intelligence. These ideas find a
platform in the media that disseminate them uncritically and resonate somewhat with
the public. I believe that the scientific community must react to this mixture of
obscurantism and buzz and propose a sober assessment of the prospects opened up
by artificial intelligence based on scientific and technical criteria.
Unfortunately, people enjoy believing in thrilling stories and imaginary dangers.
Public opinion is easily impressed by hypothetical threats, such as alien invasion or
the end of the world according to the Mayan calendar. In contrast, understanding real
risks and responding to them as a result of a calm and rational analysis always comes
too late (e.g., the foreseeable economic crisis which, nevertheless, crushed us).
Science fiction literature is full of stories of evil robots and aliens that threaten
humanity.
I quote the three basic rules of ethics which, according to the renowned science
fiction writer Isaac Asimov, robots must observe: (1) do no harm humans, (2) obey
humans, and (3) protect themselves.
Of course, the computing systems we use are not the “evil robots” of science
fiction. However, no one questions whether these systems can breach any or all of
these rules with enormous consequences—and, of course, we cannot blame them of
bad intentions, as is the case in scientific fiction stories!
82 6 Human vs. Artificial Intelligence

6.4.2 Real Risks and Challenges

Perhaps the buzz about the purported intelligence of computers conceals other, real
risks. This is where problems arise, particularly of a social and political nature,
concerning the type of social organization and the relationships they serve.

6.4.2.1 Unemployment

A danger identified long ago regarding ever-increasing automation is unemployment

for professions where the use of robots will eventually come to prevail. Thus, we will
see a gradual drop in the number of jobs in sectors such as agriculture and industry,
as well as services that can be automated. There will be employment in occupations
that require intensive creativity, for example, system programming and design, or
that require no qualifications and are not easy to systematize, for example, mail
distribution. This trend, in addition to the high unemployment rate, will broaden the
gap between well-paid jobs requiring skills and knowledge, and other manual
occupations.
Of course, some people argue that the disappearance of jobs will be balanced out
by the creation of new needs. I believe that the problem of unemployment and the
wage gap will deteriorate unless radical reforms are effected on the structure of
occupations—a subject I will not discuss in further detail.

6.4.2.2 Security, Safety, and Risk Management

The lack of security of information systems can prove extremely dangerous when the
degree of automation integration exceeds a certain threshold. It is well known that it
is not possible to achieve total protection against cyber-attacks, even for the most
vital systems. At best, we can hope to detect intruders in time. Unfortunately, for
technical and other reasons, computer systems will remain equally vulnerable in the
foreseeable future. This means that we cannot rule out disasters, particularly those
taking place during times of tension between countries. Today, cyber warfare is not
waged by individual hackers, but by well-organized businesses and even states.
A related safety risk is the interdependence of solutions offered by an increasingly
complex technical infrastructure built in an empirical manner—like the pyramids.
We know that we cannot change complex software written in old languages, such as
Fortran. Unfortunately, large computer systems are not always structured in a
modular manner like electro-mechanical systems. It is not easy to disconnect some
of their components and replace them with equivalent or even better ones without
disrupting their functioning. Such changes involve risks that are difficult to assess.
This difficulty of replacement and evolution makes reliance on certain initial
options binding for the future. For example, if you were to switch from right-handed
to left-handed traffic, the cost and risk would be enormous. Currently, Internet
6.4 Artificial Intelligence—Threats and Challenges 83

protocols do not possess the desired characteristics in terms of security and reactiv-
ity. However, initial choices commit us with regard to their fundamental
characteristics.
I have already discussed a significant difference in risk management that was
introduced by artificial intelligence and autonomous systems. There are no longer
independent state institutions that guarantee and control the quality of systems and
their safety. This responsibility is passed on to the manufacturers! The risk is
obvious, since the user safety level will be determined not by technical and trans-
parent criteria, but by an optimum between manufacturing costs and insurance costs
to cover accidents.
Unfortunately, the current massive adoption of information and telecommunica-
tions technologies is seen as a one-way street. No one is asking what kind of
technologies we should develop and why, or how to use existing technologies in
the most appropriate way. There is no public debate on the economic, social, and
political impact.
Governments and international organizations are conspicuously absent and inert.
It is as if they believe that technological progress is an end in itself. They care little
about unlawful behavior taking place on the Internet. They leave it a “free-for-all,”
considering the risks to be inevitable, as if progress entails certain unavoidable and
uncontrollable ills.
Major technology companies run campaigns with silly slogans, such as “tech for
good” or “tech for safety.” Google, Twitter, and Facebook all have sonorous mission
statements such as “don’t be evil,” “help increase the collective health, openness,
and civility of public conversation,” and “give people the power to build community
and bring the world closer together.” Of course, it would be naive to expect these
companies to care about the social problems created by innovation and the techno-
logical revolution.
Public opinion thus remains disoriented and, to a certain extent, manipulated by
voices that are neither competent nor neutral. Some exaggerate the risks, while
others promote tolerance and the take-up of technologies by downplaying the
risks. The public is ready to accept the wrong ideas and to conform.
The increasing automation of processes and services that are critical to the
seamless functioning of the economy and the organization of society leads to a
centralization of decision-making with regard to control of the cyber-world. The
problem is political in nature: the democratic control of decision-making centers for
the rational and safe use of infrastructure and services.

6.4.2.3 Technological Dependency

The prevalence of a technology that solves a host of practical problems and makes
life more comfortable means that certain skills that we developed to solve such
problems fade away. For example, very few people today know how to light a fire
using friction, a skill prehistoric people have fully mastered. Nor do people today
know how to survive in the wild or build a hut for protection. In the future, children
84 6 Human vs. Artificial Intelligence

may no longer learn the multiplication table, which was fundamental in children’s
mathematical education up to and including the twentieth century. It is reasonable
that the question of over-dependence on technology would be raised. Which basic
skills/knowledge must we absolutely maintain? The question is now being raised
even more fervently, since technology not only solves individual problems, but also
provides comprehensive solutions that imply a different way of life where it relieves
us of the burden of managing decisions by offering a plethora of services.
In order to explain the risks involved, certain people refer to the fable of the
“boiling frog” [5] placed in a pot of water. If we were to suddenly raise the water
temperature, the frog would jump out of the pot. However, if we were to gradually
raise the temperature, the frog would initially feel pleasant and would remain in the
pot until it died.
Currently, users of social media are willing to provide, in exchange for services of
dubious quality, their personal data, which is a valuable basis for controlling public
opinion. Each person’s personal preferences and information certainly have no
commercial value for them. However, the result of intensive analysis of very large
datasets is crucial for the predictability and control not only of resources and
systems, but also of markets and behaviors. Perhaps we do not understand the
importance of this information since it remains secret and solely at the disposal of
those in power and those who pay a handsome price for it.
I discern two threats to individual liberties.
1. One is the violation of privacy in order to allegedly protect societies from
infringing behavior, such as terrorism or crime. There are many plans to increase
the monitoring and control of individuals by developing appropriate technolog-
ical solutions. I would mention, for example, reputation systems that allow rating
each other in online communities in order to build trust through reputation [6].
Needless to say, it is necessary to have a regulatory framework for the use of
such tools that violate privacy and can be used to stigmatize or exclude citizens
through arbitrary processes and criteria.
2. The other threat comes from the increasing use of autonomous services and
systems in the name of efficiency. This vision is being promoted through the
Internet of Things. It envisages automation of critical resource and infrastructure
management systems without human intervention. Decision-making criteria may
be so complex that they exceed human possibilities of understanding and control.
We may thus reach a system of techno-tyranny, all the more because automation
enables the concentration of decision-making at ever fewer centers.
Therefore, the real risk is that computer knowledge will be used uncontrollably
to make decisions, replacing people in critical processes.
The proper and rational use of artificial intelligence and autonomous systems
depends on two factors.
1. The first factor is the possibility of assessing, based on objective criteria, whether
we can trust the knowledge generated by computers. This is currently the subject
of research which, one is to hope, will be able to give us the necessary results.
References 85

We are trying to develop “explainable” artificial intelligence using techniques

that allow us to understand and, to some extent, control the behavior of systems.
I believe that we need a new method of assessing knowledge, bridging the gap
between scientific knowledge and the empirical knowledge of artificial
intelligence.
2. The second factor is the vigilance and sense of political responsibility of society
as a whole. When computers use knowledge to make critical decisions, we need
to be sure that this knowledge is safe and neutral. The safety of autonomous
systems must be certified by independent authorities and not be left to those who
actually develop them. Here, the existence of a legislative and regulatory frame-
work and the role of institutions in risk management, which will be discussed in
subsequent chapters (see Sect. 7.3.2), could come into play.

References

1. https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow
2. https://en.wikipedia.org/wiki/History_of_artificial_intelligence
3. https://en.wikipedia.org/wiki/Correlation_does_not_imply_causation
4. https://en.wikipedia.org/wiki/Technological_singularity
5. https://en.wikipedia.org/wiki/Boiling_frog
6. https://en.wikipedia.org/wiki/Reputation_system
 Part III
Consciousness and Society

We seek to analyze the functioning of individual consciousness and to discuss its

impact on societal organization, in the light of the previous gnoseological
perspective.
• We explain the functioning of consciousness as an autonomous system that
manages short- and long-term goals based on value criteria and accumulated
knowledge.
• We discuss how values are shaped in society and how the composition of each
individual’s subjective experience takes on an objective dimension, which can be
studied as a social phenomenon.
• We present an analysis of the role of institutions in shaping and maintaining a
common scale of values for strengthening social cohesion, as well as a discussion
on the principles of democracy.
Chapter 7
Consciousness

7.1 Our “Big Bang”—Language and Consciousness

7.1.1 Language and the Boundaries of Knowledge

Of all the seemingly insurmountable problems we face, one looms the largest and
most important. How did cognition and language appear through evolution? How
was consciousness formed? Questions concerning cosmogony or the evolution of
the species appear trivial in comparison to the question, which sadly has been
hitherto subject to fragmented research in various disciplines (biology, psychology,
informatics, linguistics, anthropology, etc.). How are abstract concepts created
through experiences? How do we comprehend by linking phenomena to analogies
and metaphors? How are the boundaries of knowledge and understanding of the
world demarcated? How do we ascribe meaning to symbols? How do we create?
How does awareness function as a system that combines wishes, motives, and will?
How do we choose goals and act on them?
Language defines the coordinates of cognition: being able to mean what you say
and to clearly state what you mean. What you feel is more than what you can
express. What occurs is more than what you can describe. Only what can be placed
in the semantic relationships of language can be understood. By language, I mean
not just natural language, but also internal and ineffable language, as well as any
manner of structure that can be interpreted and carry information.
Important research programs are focused on studying the functions of the human
brain. I do not question the importance of such research, which will teach us much
that is useful about the functions of this wondrous “processor.” However, I do not
think that, due to their very nature, they are able to answer the questions at issue.
Mental phenomena must be studied on a different scale and cannot solely involve
analyzing the signals and neural circuits of the brain.
Many will claim that these problems can only be the subject of philosophical
inquiry, and that a strict scientific approach is impossible or even misplaced. I would

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 89
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_7
90 7 Consciousness

counter that such an assertion must be supported by arguments as to what exactly a

scientific approach is and what its limits are. However, this is part of the questions
we have raised.
In the past, problems, which were the subject of philosophical inquiry, entered the
sphere of scientific analysis. Mental phenomena complement physical phenomena.
They are the other side of reality. Attempting to understand them through scientific
methods is a worthwhile, imperative endeavor. Such an undertaking requires broad
interdisciplinary cooperation, and I believe that the contribution of disciplines, such
as informatics, biology, and medicine, could be decisive.
In the future, I would like to see the same generous funds and mobilization of
large research teams dedicated to exploring space being used to set up programs to
explore the mind as a tool, which on the one hand, ascribes meaning to the world
and, on the other hand, is the starting point for decisions on action and creativity.
Perhaps then we might be able one day to penetrate the mystery of the “Big Bang of
consciousness,” and the theories of physical cosmology might be complemented by
other, parallel ones that would illuminate our unexplored selves and bring us slightly
higher on the wonderful scale of self-knowledge.
To what extent can we formalize natural languages in order to be able to delve
deeper into their concepts and relations? This remains an open problem despite
advances in mathematics and logic. Formalization requires a set of core independent
concepts to which all others refer. When concepts are not independent, they should
be linked through semantic relationships, as is the case in formalized languages.
What makes formalizing natural languages very difficult is the lack of adequate
semantic models, as explained in Sect. 6.1.2.
One major advantage of humans is that natural languages are discrete. Phrases
consist of words, and words consist of syllables.
How would we understand the world if language was not discrete, but continu-
ous? What if we talked in whistles, like birds? With no consonants or vowels? I am
not referring to well-known whistling languages that simply translate discrete natural
languages. I am imagining languages where every concept or phrase would be a
whistle, which, in my opinion, would severely restrict their expressiveness. This
would depend on our acoustic capabilities—that is, the frequency interval of the
acoustic spectrum that we could perceive.
If that were the case, I cannot imagine how we would define mathematics and
algorithms. Could an algorithm be a transformation from one continuous signal to
another? I believe this passage from the continuous to the discrete is integrally linked
to the development of conscious thinking.

7.1.2 The Mind–Brain Relationship

Another interesting question is the mind–brain relationship. Can all mental processes
be understood by exclusively studying the brain? Let me express my doubts by
proposing an analogy. Let us suppose that your computer evaluates a mathematical
7.2 The Mind as a Computing System 91

function and that an engineer can observe the state changes of all its circuits during
the computation. Is it possible, from observation alone, to determine the function
whose code the computer executed? Without going into technical arguments, the
answer is that this is an extremely difficult problem. I believe that we will not be able
to understand mental functions by exclusively studying the brain as a physical and
information processing system (see also the discussion on emergent properties in
Sect. 4.1.4).
We currently have techniques at our disposal to study how concepts are
represented in the brain as a pattern of electrical signals. One interesting question
is whether the way in which data are represented and generally organized is the same
for all individuals, and especially whether it depends on the language we speak.
Recent thought transmission experiments using computers (Brain Computer
Interaction) tend to demonstrate that concepts have a canonical representation in
the brain. In other words, the same concept, regardless of language, is represented
through the same combination of signals and states (patterns). If this is the case, it
should raise concerns for those who believe that we are the result of random
evolution. This would mean that, although we do not speak the same languages,
there are common origins in terms of the representation of concepts. If it is exper-
imentally proven that the representation of a concept such as “door,” “window”—or
much more so “freedom”—has common characteristics regardless of language, this
will be a very interesting result. It may even shed light on the problem of the
existence of language families.
The theory of the existence of a Proto-Indo-European language that developed in
a certain region and then spread with differentiations across almost all of Eurasia
does not seem so convincing to me. The solution to the mystery should be sought
elsewhere. Research into the representation of concepts in the brain may help us to
reach some interesting conclusions.

7.2 The Mind as a Computing System

I have written this text as a computer scientist and engineer, relying particularly on
models and ideas that I have developed for autonomous systems that I have been
studying in recent years. As a non-expert, I have no claim to the “scientific veracity”
of my views. It is rather a conjecture into how the human mind could function as an
information system based on my knowledge and what I have understood by reading
relevant texts from psychologists, neuroscientists, and philosophers. I would like to
stress that these texts contain a fairly wide range of views and are characterized by an
absence of organization of concepts and the relationships between them. As a reader,
I found very little that was actually possible to digest.
What I tried to do is to imagine an architecture where the various elements of
consciousness function in a technically feasible way, insisting more on relationships
between concepts that determine the flow of information and knowledge, without
92 7 Consciousness

this meaning anything about the model’s accuracy regarding what might actually be
occurring.
I once wrote that scientific knowledge is nothing more than a set of myths, which
have been tested in practice with systematic experimentation and validation. Such
myths make airplanes fly, keep bridges and buildings standing, and illuminate
continents when the night falls. However, they are but a poor approximation of
physical reality, to the extent that we can observe and understand it.

7.2.1 Value Systems and Regulatory Frameworks

Based on which mechanisms and criteria does the mind exercise freedom of choice,
if, of course, it can actually choose? If so, how does it choose between many diverse
needs, both material and intellectual, when it has limited resources?
One simple way to understand these mechanisms is by defining a common
benchmark.
Let us consider that each individual has a value system using scales of values to
determine for each action, the units of value required for it to be carried out and,
possibly, the units of value it produces. Such an idea has been successfully
implemented in many fields of knowledge. It is the idea of the “invisible hand”
that Adam Smith imagined for the economy. It is the idea of the resource for
computing systems, which I explained in Sect. 5.2.2. The laws of physics can also
be regarded as a value game. If I want to move and go from one place to another, I
need resources, energy in this case, which depends on my mass and the relationship
between the coordinates of the positions.
Of course, it might seem strange to use numerical values, considering that any
kind of human action has a “value balance.” For example, how would I measure the
intensity of the feeling of hunger, which I can “zero” with a meal that costs 15 euros?
However, if we wish to find a functional scheme that correlates the various criteria
entering the consciousness game, we have to accept a “common denominator” and
“equivalence relationships” in order to compare needs against each other and match
them with the resources needed to satisfy them.
I will also consider that any social organization in a place and time is based on a
common value system that emerges as the “synthesis” of the individual value systems
of its elements. Similarly, this system uses common scales of values for evaluating
the actions of individuals, enabling us to understand how societies and individuals
function as dynamic systems, more or less coherently. This is achieved thanks to the
existence of a common framework for evaluating actions, encouraging and reward-
ing beneficial actions, while discouraging and penalizing actions that harm society as
a whole. Hence, social cohesion and the synergy of individuals to achieve common
goals are reinforced.
The values of common scales and their nature change according to the field of
action, i.e., economic, political, legal, educational, military, gnoseological, moral,
religious, and esthetic. We sometimes relate values to principles because they
7.2 The Mind as a Computing System 93

indirectly define what is important or even provide a basis for comparison in order to
set priorities between our actions and appropriate incentives to implement them.
For example, the economic value system concerns monetary values, which are
regulated in order to help the orderly production/exchange/use of goods and ser-
vices. The incentive in this case is to generate profit.
Political values encourage good governance by making timely and key decisions
for societal prosperity and safety. Legal values are used to administer justice by
preventing the violation of laws through penalization. Educational values, such as
excellence, and enthusiasm for learning and creativity, are intended to support the
transmission of knowledge and to prepare citizens capable of contributing to and
integrating into society. Military values foster a spirit of readiness and self-sacrifice
to safeguard a country’s security.
Furthermore, moral values define what is good in our behavior, while
gnoseological values promote sound knowledge and information. Religious values
aim to strengthen relations with the divine through a set of rules and practices.
Finally, the scale of esthetic values sets criteria for what is beautiful or pleasant, such
as a work of art, an environment, or an experience.
The above list is not exhaustive and is provided simply in order to explain the
concepts introduced.
I will assume that we use scales of values to correlate the values that each action
consumes and generates. Positive value means profit, recognition, satisfaction, for
example, when I pass a test, or when I earn money. Negative value means loss of
money, a penalty or shame, for example, when I am arrested because I violated road
traffic rules or when I am seen mistreating an animal.
A specific category of actions with a negative value are prohibitive actions with a
dissuasive cost for the person who commits them, conferring upon such actions a
deterrent effect. These include criminal offences, such as killing or robbing some-
one, which are punishable by law, as well as actions, which, albeit not punishable,
involve excessive risks, such as spending recklessly and racking up debt.
Scales of value also define the mandatory actions that we should reasonably be
required to take because doing nothing would entail prohibitive value costs. Thus,
you have a legal and moral obligation to help someone in danger, to respect road
traffic rules, etc.
Note that prohibition and obligation are not independent concepts: a prohibition
could be expressed as an “obligation not to.” We can therefore say that “it is
forbidden not to comply with road traffic rules.” However, in practice we use both
concepts for the sake of clarity.
Actions that are not mandatory or prohibited are optional. They involve a
relationship of reciprocity. The criterion for their execution is whether their perfor-
mance is considered to be advantageous. Optional actions define the framework for
individual freedom where public conscience is exercised with a view toward the
common good and societal prosperity.
It should be noted here—and will be made even clearer later—that, given the
heterogeneity of resources (time, money, physical and intangible resources) and
quality criteria, it is not simple for one to decide what is in one’s interest and what is
94 7 Consciousness

not. In taking a certain action, I may use money to gain time and/or social recogni-
tion that cannot be quantified.
Furthermore, an action can also affect more than one field. An economic action
can have consequences that are not only economic but also social and moral. On the
other hand, a morally motivated action, for example, charity, can have economic
effects.
Although the values of the various fields are not comparable, when we make a
choice, whether consciously or automatically, we are using an equivalence relation-
ship. For example, when you decide to pay 200 euros in order to buy a pair of shoes,
you take into account your financial capability, as well as other esthetic and social
values that do not arise from your need for footwear.
Often, one way to promote values that concern non-material goods, such as road
safety or patriotism, is to encourage practices or to apply regulatory frameworks that
lay down mandatory or prohibitive rules. Thus, road traffic rules promote road safety
and are part of the statutory regulatory framework. Right-hand priority rules or speed
limits restrict the behavior of drivers in order to enhance safety. Therefore, the
application of a traffic rules system allows for something that cannot be practically
achieved by abstractly explaining to the general public that road safety is an
important value. That is why social organization around a value system is achieved
through the application of regulatory frameworks. There are countless examples in
every field.
If virtue is considered to be a supreme moral value, ethical rules help you to be
virtuous either by preventing you from acting immorally or by indicating how you
can be virtuous in certain circumstances, for example, by encouraging charity or
mutual aid.
In education, there are rules that encourage and recognize excellence, for exam-
ple, students who excel are valedictorians at graduation ceremonies, are awarded
scholarships or enjoy other advantages. The above demonstrate the importance of
regulatory frameworks for social organization, which we will further discuss in
Chap. 8.
In order to study the behavior of individuals, I will assume that each person has an
individual value system that largely reflects the values of a common value system,
especially those that we cannot disregard without cost, for example, financial or
legal. The individual value system also includes rules regarding values that charac-
terize actions of a personal nature, such as in the field of ethics, religion, or esthetics.

7.2.2 The Components of Consciousness

It is not easy to define consciousness, as the concept is quite complex and has not
been fully explored. I will begin with a simple definition, which will be enriched in
this chapter, by saying that consciousness is the ability of humans to understand the
world and act in response to internal and external stimuli. Consciousness manifests
itself as an interactive game between the mind and its environment. On the one hand,
7.2 The Mind as a Computing System 95

Fig. 7.1 Architecture of a mental system with its key functions

the mind plans and carries out actions to meet the physical and non-physical needs of
the individual. On the other hand, the environment determines the economic, social,
and physical conditions in which individual freedom is exercised.
In this game, one’s individual scale of values plays a central role in exercising
freedom of choice and achieving one’s goals.
I am not seeking to explain exactly what consciousness is and how it actually
works. My aim is to define, as simply as possible, the architecture of an information
system, which I call a mental system and whose behavior approximates that of
conscious thinking. I use the term “approximates” because I focus on decision-
making issues and I ignore others, perhaps more important, that concern self-
learning and self-awareness, as well as cooperation between the conscious and the
subconscious.
I assume that the mental system is an autonomous system (Fig. 7.1), similar to the
one described in Sect. 6.3.2, suitably enriched to take into account the particularities
of humans.
The mental system interacts with the external environment by obtaining sensory
information and performing actions. In the case of humans, the external environment
is the entirety of external phenomena, both physical and social. The internal envi-
ronment is the body with the senses and the motion and speech organs. By using the
functions of perception and reflection, the mind builds a model of the environment,
as I have already explained.
One particular feature of human beings is the ability to self-learn. Humans can
create new goals to meet different kinds of physical and non-physical needs. This
admirable adaptation mechanism is based on a capability for introspection, which
96 7 Consciousness

interprets the states of the environment model and produces feelings, which I will
discuss immediately below. For the moment, I will say that feelings are mental
states, which are accompanied by emotions and are directly linked to needs. When
the intensity of a feeling exceeds a threshold, the will to satisfy the need arises by
setting the appropriate goals, to be managed by the decision-making function. Goals
are determined by taking account of the individual scale of values, the state of the
environment and accumulated knowledge regarding the environment, expressed in
the form of general rules.
I consider that this architecture of the mental system, which I will explain in
further detail, consists of three components dealing with: (1) understanding and
building a model of the environment with which we interact; (2) the emergence of
needs and the will to satisfy them by setting the relevant goals; and (3) decision-
making concerning the management of goals, and their planning and realization.
In conclusion, I would point out that Plato also distinguishes three key parts in the
human soul, which are not unrelated to the functions of the model: the appetitive part
(epithumetikon), the spirited part (thumoeides), and the rational part (logistikon). All
the desires of a person, as expressed in his or her interaction with the environment,
emanate from the appetitive part. The will to satisfy needs by taking account of a set
of values emanates from the spirited part. Finally, the rational part concerns the
coordination of the other parts and the management of knowledge.

7.2.3 The Mental System

7.2.3.1 Feelings and Needs

Through the process of introspection, the mind creates mental states that describe the
intensity of feelings. I will not attempt to define the concept precisely, as there is no
agreement among experts. I will simply say that a feeling is a subjective conscious
experience, combining mental conditions and psychosomatic reactions.
As mental states, feelings have intensity and a “sign.” Generally, a feeling is
pleasing when it satisfies criteria concerning the entire spectrum of quality of life,
from comfort to survival. The criteria are subjective but are influenced by objective
factors. Therefore, for some reason I may be feeling cold, even if the temperature is
not low. The intensity of feelings is different for different people under the same
environment conditions.
Feelings are linked to the satisfaction of needs: biological, safety, social, self-
esteem and self-fulfillment, intellectual, etc.
Biological needs are expressed through feelings of thirst, hunger, sleep, fatigue,
and others concerning the satisfaction of physical needs. Safety needs concern the
stability of work relationships, adequacy of financial resources, living conditions,
social security coverage, etc. The related feelings, such as fear or anxiety, stem from
our assessment that we are facing a risk, based on the state of the environment model
and our knowledge from previous experiences.
7.2 The Mind as a Computing System 97

Satisfying social needs manifests itself through positive feelings stemming from
the existence of stable and substantive esthetic, friendly, and familial relations.
Failure to satisfy such needs entails feelings of rejection and abandonment.
Self-esteem and self-fulfillment needs are expressed through positive feelings,
such as self-confidence, professional recognition, and feelings of excellence due to a
position in a hierarchy.
Finally, intellectual needs are grounded in non-material values and are expressed
through feelings such as compassion, solidarity, temperance, and honesty.
Experts suggest detailed classifications of feelings and the related emotions. I am
not competent to speak on the matter. The above list is not exhaustive and is
provided simply for argument’s sake.
I highlighted the distinction between positive and negative feelings. Another
interesting distinction is between passive and active feelings.
Passive feelings are those that do not urge a response. They stem from a situation
that we are experiencing and that does not give rise to other feelings. Such feelings
are fear, sadness, shame, and pain.
Fear, for example, stems from the assessment that we are in a dangerous situation.
This may be objectively justified by a reasonable analysis of the environment state. If
the fear we feel is exaggerated, then we are talking about a phobia. When travelling
by airplane, there is always a minimal risk, which is assessed according to technical
criteria. The behavior of those on board, reacting as if there is a high likelihood that
the worst will come to pass at any moment, can be considered a phobia.
The opposite feeling, fearlessness, is active and means a rash underestimation of
risk. Examples include participating in an extreme sport without taking the necessary
precautions, or indiscriminately spending money and reducing our financial
resources beyond the threshold necessary to survive.
Active feelings suggest situations involving alertness or stimulation of mind, such
as enthusiasm, passion, interest, ecstasy, and responsibility.
For example, a feeling of responsibility means constant vigilance to weigh the
pros and cons of our actions and to take their possible consequences into account.
Being responsible does not just mean doing the right thing according to certain
criteria, but also acting with all the care necessary to achieve the best result. It is easy
to prohibit children from doing something they should not do, but it is more
responsible to convince them to do the right thing through dialog.
An irresponsible person will take a cavalier attitude of over-optimism regarding
the outcome of their actions, perhaps taking risks that could lead to dangerous
situations. A responsible person will endeavor to weigh the risk by using their
knowledge and objective criteria as much as possible. They can thus justify their
actions as a result of an analysis that takes various criteria into account, ranking them
hierarchically.
A responsible person has a reasonable attitude toward risks. One interesting case
is heroism, when a dutiful person overcomes fear. Heroes defy fear not because they
do not properly weigh the risk, but because they are driven by a sense of duty. They
consciously carry out a dangerous action, considering the defense of intellectual
values superior to their own personal safety.
98 7 Consciousness

How do feelings relate to needs? Having assumed that every feeling relates to a
need, I believe a need arises when the intensity of a related feeling exceeds a certain
threshold.
I will represent the intensity of feelings with variables defined in an interval of
values between a lower and an upper bound. Conventionally, positive values express
satisfaction, negative values express the opposite, while 0 is an indifferent state. We
have pleasant feelings (positive value), unpleasant feelings (negative value), or even
indifferent feelings (zero value) by assessing the states of the environment. Thus, I
may feel indifferent toward sports news, the fact that whales are being decimated en
masse or that climate change is looming over us.
For example, I can use a scale between +5 and 5 to express the intensity of the
feeling of hunger. Conventionally, 5 means I am absolutely sated, +5 means I have
an absolute need for food, and 0 reflects an indifferent state. Of course, every person
has his or her own personal tolerances.
The intensity of feelings on an individual scale of values plays an important role
in the choice of actions and goals.
I believe that for every feeling there is an optimal value which I will call a point of
equilibrium. It is the most painless state for negative feelings and the most pleasant
state for positive feelings. Thus, I can imagine that the mind chooses the actions
carried out, so as to be close to the point of equilibrium for all feelings. Such a
situation is difficult to achieve. This is because, when I try to change the intensity of
a feeling, for example, to restore the feeling of hunger back to 0, the point of
equilibrium, I may disrupt the intensity (state value) of other feelings.
I have said that a need arises when the value of a feeling exceeds certain limits on
the value scale. If, for example, the intensity of the feeling of hunger approaches
critical values, then the system inquires whether and how this need can be satisfied
by choosing appropriate goals. In other words, a goal concerns the attainability of
states of the model that satisfy a condition—in this case, access to food. The
correspondence between needs and goals is knowledge developed through the
self-learning function based on experiences. That is, I know from experiences how
and where can I find food if I am feeling hungry. Goals are selected through the
decision-making functions of the mental system, as explained in what follows.

7.2.3.2 States and Actions of the Mental System

The mental system is a system with states and actions, such as the one we presented
in Sect. 5.2.2. It has a very large set of states defined by the values of three types of
variables.
1. External variables describe the state of the external environment.
For the physical environment, variables include those that determine location
and time, weather and environmental conditions such as temperature, pollution,
and noise.
7.2 The Mind as a Computing System 99

The economic environment is characterized by a set of variables that have a

particular impact on the assessment of one’s personal financial situation in order
to take actions of an economic nature. These concern the income, debts, taxes that
we pay, but also other, more general parameters, such as the level of interest rates
or property prices.
Other variables are those that affect decisions on all types of social services,
such as education, health, and protection, for example, the cost of medicine and
hospitalization, and the level of social benefits.
2. The variables of the internal environment characterize our body’s state according
to physiological parameters such as temperature, heart rate, and blood pressure.
3. Finally, the model also possesses a variable for every type of feeling that
describes its intensity using values of the respective value system.
Furthermore, the system has two types of actions:
1. Actions controllable by the mental system, which may decide whether to carry
them out when their preconditions are met, for example, withdrawing money,
travelling, and shopping.
2. Actions non-controllable by the mental system, carried out in its environment by
other persons or with natural causes. Actions by external factors, such as the
expiry of a deadline or a change in the weather, can alter the state of the mental
system regardless of our intention. There are also internal non-controllable
actions, which are automatically carried out by internal mechanisms and lead,
for example, to feelings of hunger, thirst, etc.
When we perform a controllable action from a state, the mental system transitions
to a new state, changing the state variables according to the amount of values that the
action creates or consumes, as defined by the individual value scale. This action, as
we have said, cannot ignore the common scale of values, particularly with regard to
changes in the variables of the external environment.
As in Sect. 5.2.2, an action is defined by the preconditions for its execution at a
state of the model and by its result when performed.
1. The preconditions for an action to be executed at a state are qualitative or
quantitative requirements that the state must satisfy. The quantitative require-
ments relate to the availability of a quantity of physical resources and/or intan-
gible goods measured on the applicable value scale. Resources are committed
and/or consumed to render the action feasible. For example, a precondition to go
shopping is to have the necessary funds available. A precondition to dare to go
skydiving is to be in the right physical and mental state to overcome the feeling of
fear. Qualitative preconditions are assertions about the system state that either
hold true or not. For example, in order to read an e-mail message, you must have
access to a network and an electronic device.
2. The result of the execution of an action is to move from the present state to a new
state by changing values of state variables, in particular by consuming and
creating values. For example, when cooking baked pasta, I use resources, namely
basic foodstuffs and energy, and I produce, for example, 10 portions of baked
100 7 Consciousness

pasta. When I sleep, I consume “active” time and proportionately increase

restfulness.
Not only controllable actions but also non-controllable actions can affect our
mental state. This is because, through our senses, information regarding a change,
for example, in room temperature, in bank interest rates, and in a building’s plot
ratio can affect our feelings, just like the satisfaction or non-satisfaction of a need
for heat results in optimism or frustration.
By extension, an action stemming from altruistic feelings can also change
external variables. Helping an injured or disadvantaged person is an action that
increases my self-esteem, but also involves interaction and therefore changes to
the external world.
It is interesting to understand how material and non-material values are linked
when we perform an action, such as “having lunch at a restaurant.” There are
preconditions such as “being at a restaurant,” “having the necessary funds,” and
“the feeling of hunger exceeds a certain value.” Immediately after the meal has
ended, I will still be at the restaurant, but the funds available will be fewer,
depending on the amount spent on the meal, and I probably would be quite sated
(satisfaction of appetitive needs). Here I am using the value scale for the feeling of
hunger. We are not interested in exactly how it is determined and how it is graded,
which may vary from person to person. However, such a concept is necessary in
order to study the effects of hunger on the behavior of individuals and their
environment.

7.2.3.3 Selection of Actions—Conflicts

We already discussed conflicting actions in Sect. 5.2.2: when two or more actions
from one state are possible and realizing one can cancel out the preconditions for
realizing the other. There are many everyday examples of conflict situations, which
the mind must resolve in the best possible way between the controllable actions of
the model from a given state. First, the choice is made between the actions whose
preconditions are valid. We know that every action can consume and release values.
Therefore, a simple selection criterion is to choose an action whose balance of values
is the “most positive” one. For example, I would prefer to use existing financial
resources in an investment with a high yield rather than buy an apartment or travel
around the world. Of course, the problem of comparing dissimilar values arises, but
it is not as theoretical as it seems, as we address it every day. Equivalences, even if
non-consciously, are applied in order to compare different value scales.
Therefore, you decide to perform charity because the moral satisfaction that arises
is more important than the financial cost it entails. The values of material goods are
transformed through the market into monetary values—but there are also immaterial
goods, such as freedom, justice, and knowledge, which should be estimated on a
value scale in order to resolve conflict situations.
7.2 The Mind as a Computing System 101

Individuals and societies must choose between non-comparable values. How

much can one exchange one’s liberties for material goods and comfort of life?
There are many such questions, and the answers we give them at a given time
depend on the way values are hierarchically ranked, as shaped less in the form of
written rules and more in our conscience.
Another criterion for choosing an action is the criticality of a situation. For
example, lack of a resource can have negative effects, such as an impasse when
the intensity of a feeling is far from its point of equilibrium. Thus, if my funds are
dramatically reduced, then the choice of “emigrating to work abroad” can become a
priority over other choices, even if it entails a high emotional cost. Similarly, if I am
very tired or hungry, the actions that will bring me back to the point of equilibrium
take priority over others that may have a more positive balance of values.
It is worth noting here that the results of actions do not always have the same sign
for both the individual and society as a whole. One characteristic case is that of a
criminal, whose action may be harmful to others, while its outcome is positive for
him or her.
The behavior of an upstanding citizen is such that they do not take actions that are
harmful to others, even when they suffer harm themselves. There is, of course, also
the case of the fools, who may indiscriminately take actions that harm both them-
selves and others.
This simple model that views the conscious execution of an action as the result of
a decision, weighing the pros and cons of prerequisites and results, enables us to
understand how differences in people’s value systems result in differentiation in their
behaviors.
This differentiation is what makes a person stand out as selfless, selfish, gourmet,
heroic, rude, etc. A selfless person will take an action for the common good, even if
they stand to take a loss. A selfish person would not accept any action that would
entail a personal cost, no matter how favorable it would be for others. A gourmet
may place culinary values higher than monetary values in comparison to the average
person. A hero would take action for the common good even if they entail a
prohibitive cost. A rude person would act in a way that does not respect established
rules of behavior toward others.
In conclusion, I would say that the mental system aims at the harmonious
operation and oversight of the whole, while maintaining a balance between satisfy-
ing material needs and value requirements. This cooperation/struggle between desire
and spirit or mood, under the arbitration of informed thinking, results in highlighting
short- or long-term goals.
In what follows, I will try to explain how the elements of consciousness work
together to achieve goals.
102 7 Consciousness

7.3 Goal Management and Planning

7.3.1 Strategies and Tactics for Achieving Goals

When there is a need that cannot be satisfied through immediate action by the current
state of the mental system, the mind must identify possible goals, i.e., attainable
model states through which it can be satisfied. We already said in Sect. 6.3.2 that
there are positive and negative goals. Satisfying a need can be achieved through
different positive goals to different degrees. However, every positive goal also
results in other negative ones. Negative goals are linked to the cost of risk, as they
concern avoiding dangerous situations entailing considerable value cost. Risk
depends on the possibility of finding ourselves in such situations due to various
uncertainties.
For example, an economic action involves transactions with other persons who
may not be solvent or who are affected by political and social changes. When we
decide to partake in a dangerous sport we enjoy, we take the relevant risk into
account when calculating the cost. This also applies when we travel or generally
change our state within the environment. There is also the risk of facing legal
consequences when we act unlawfully. Finally, there is the risk that arises from
satisfying needs when the action may endanger our physical integrity.
Therefore, when assessing the cost of achieving a positive goal, we must also
incorporate the risk of the associated negative goals. This is an extremely complex
problem, because a tactic is not a simple sequence of actions, but a set of sequences
of actions depending on the reactions of the environment. Thus, the cost estimate
must be graduated taking into account the “worst case scenario” and the resources
available.

7.3.1.1 Strategies

In a given state of the mental system, a set of needs and corresponding goals emerge
in the mental system, which the mind initially ranks depending on their importance
and criticality. As is the case with actions, a goal becomes critical when failure to
satisfy the need in time can lead us to catastrophic situations with huge value costs,
whether material (money, health, and integrity) or immaterial (deprivation of liberty,
moral cost). Thus, the mind tries to achieve critical goals in order of priority,
planning appropriate tactics and mobilizing the necessary resources.
Non-critical goals may also be classified according to the cost of not achieving
them in time. The mind makes comparisons between similar goals, for example, to
go to a restaurant or to the movies on a given night, both needs arising from the
desire to “blow steam off,” taking into account all the knowledge relevant to their
achievement. It will choose from a variety of economic and technical criteria (how
easy it is to visit each locale from one’s home) and personal preferences and
memories from any past experiences. When we have satisfied one of the two
7.3 Goal Management and Planning 103

goals, on the way back home we make a conscious or even automatic assessment of
whether the choice made was good (whether it had the results anticipated). If not,
then the self-learning function will change certain model parameters that will affect
the relevant choices in subsequent judgments.
The hierarchization of goals into short- and long-term goals is taken into account
together with their prioritization. Short-term goals are those that we manage every
day and have a sense of urgency. There are also exceptional short-term goals that are
usually given higher priority.
Long-term goals must be achieved over time and, to this end, the mind must carry
out a more thorough analysis in order to define a strategy, taking into account the
resources available and external factors. Their management requires special dili-
gence and perseverance because we often tend to be diverted by short-term goals,
even if they are not important.
We can consider the development of strategies to manage goals as a solution to an
optimization problem, taking into account constraints related to the resources avail-
able. We will choose a set of goals whose timely achievement is possible using the
resources available and which are expected to best satisfy our needs. One inherent
difficulty in this process is the lack of predictability of dynamic changes in the
environment. New goals may emerge or existing ones may need to be modified. The
more predictable the environment, the more successful our goal management can
be. This is perfectly understandable in the field of economics, where stability and
predictability of trade allow for proper resource management. Lack of predictability
implies a permanent commitment of resources to respond to emergencies. That is
why social and political stability are important factors for economic growth and
prosperity.

7.3.1.2 Tactics

When a goal is selected, the mind calculates a tactic to achieve it. The tactic consists
of planning sequences of controllable actions that may lead to achieving the goal
depending on uncontrollable environment reactions. Here lies the complexity of
calculating tactics. A tactic is not a sequence of actions, but a set of action sequences
that can be represented by a tree structure. When we perform an action that changes
the state of the environment (node of the tree), the next action is chosen in response
to the reactions of the environment (branches of the tree from the node). When the
environment offers a large number of reactions—as is often the case in practice—it is
impossible to make a static estimate of tactics to achieve a goal from a given state. A
tactic is dynamically calculated over a time horizon and updated from time to time
until the goal has been achieved.
Thus, calculating tactics appears to be a game that we play against the environ-
ment, both external and internal. The difference, though, is that in this game the rules
are not clearly defined, particularly regarding the intensity and the timing in the
exchange of “blows” between opponents.
104 7 Consciousness

Identifying successful tactics is not easy and requires not only knowledge, but
also proper knowledge management in order to both assess situations and find
solutions that keep the goal accessible.
It is sometimes greatly tempting to take an action which brings us very close to
achieving the goal, but at the same time places us in a situation where it is impossible
to go further—and so we definitely lose the game. Examples could be making an
investment by taking out a loan that I must start paying off before I reap the return on
my investment or running a race and exerting all my energy to remain first at the
outset, which I cannot keep up all the way to the finish line.
I must point out that, in practical terms, the effectiveness of the decision-making
mechanism depends on the quantity and quality of knowledge, but above all on the
ability of the mind to effectively combine knowledge. This is what I call meta-
knowledge or wisdom: a person can have a wealth of knowledge to answer basic
questions but be incapable of combining them creatively to achieve a goal. There are
well-read persons who are “living encyclopedias” yet are inadequate in managing
and achieving goals—and there are people with a moderate education with outstand-
ing judgment and inventiveness in difficult situations.

7.3.2 Safety and Liberty

“Those who would give up essential liberty to purchase a little temporary safety
deserve neither liberty nor safety,” said Benjamin Franklin. I agree with this phrase,
which has a lot of truth in it and is a fine defense of fundamental liberties.
We must not ignore the fact that liberty and safety are often opposing concepts.
When one grows, the other diminishes. The issue is which liberties we are prepared
to sacrifice and to what extent in exchange for our safety. It is up to each person to
strike the right balance between liberty and safety, based on their values and the
estimated degree of risk.
Each person must exploit his or her degrees of liberty to achieve a goal effectively
without jeopardizing his or her safety.
We have already discussed critical goals, which if not achieved, result in cata-
strophic situations. However, we face the daily risk of such situations if we do not
plan ahead so that our actions are safe.
An action is not just safe if its result does not directly lead to a catastrophic state.
Additionally, after its execution, it should not be possible for the environment to lead
us, through non-controllable actions, to a catastrophic state without us being able to
prevent it. For example, if we leave the house without locking the door, we are
creating a potentially unsafe situation, in the sense that a person can break in and
burglarize our home (catastrophic state).
Technically speaking, safety is described by conditions on variables of the mental
system that characterize safe states. Examples would include an annual income of
over 10,000 euros and a body temperature of under 40  C. Non-satisfaction of a
safety condition can lead certain actions to deadlocks. If a deadlock is permanent,
7.3 Goal Management and Planning 105

this entails a significant decrease in the degrees of liberty in the management of

goals.
One special case of safety conditions are those subject to time constraints. In
theory, if we took no initiative for action, our state would not change were it not for
the factor of time. In a sense, time acts as our “adversary.” Over the course of time,
even during a period of complete inertia, we have to eat and drink, pay our rent, go to
sleep. Thus, there are conditions determined by critical deadlines, which if not met,
can lead to an adverse result.
Complying with safety conditions generally results in reducing degrees of liberty
and, therefore, effectiveness in achieving goals. If I wish to advance too quickly, I
might disregard or fail to properly assess certain risks. For example, I might not wait
at a street crossing in an effort to reach my destination faster and thus expose myself
to the risk of an accident. It is a well-known fact that if you do nothing you reduce the
likelihood of making mistakes, but you also do not achieve anything positive. A
prudent person knows how to handle freedom of choice, assess the risks involved
and identify tactics that provide for safe actions.

7.3.3 Free Will

Deadlocks are a problem facing people, not the physical world. Clocks do not stop
the way the human mind stops when it is unable to find ways out of difficult
situations. Humans are doomed to try to find the resources that will enable them to
achieve goals. In a sense, it is the price we have to pay for freedom of choice.
We have said that social organization is based on two types of interaction:
cooperation and conflict. Cooperation is the coordinated execution of actions by
more than one person in order to achieve a common goal. Conflict is the result of
competition in order to obtain resources (see Sect. 5.2.2). Conflict may be internal or
it may arise between individuals. Human societies differ greatly from insect socie-
ties, where cooperation is hard-coded into their genetics and conflict phenomena are
rare. I have often wondered why evolution did not stop in societies organized solely
based on automatic processes, such as those of insects.
I have explained that one theoretically interesting case is a conflict-free state. This
is the case when there is an over-abundance of resources, and all possible actions can
be realized independently from each other. Air, for example, is a critical resource for
life. However, under normal conditions, no conflict arises since each of us can
breathe air independently of each other. The same holds true for water and fuel
although in this case their consumption entails a proportionate consumption of
monetary resources.
Conflict-free systems are characterized by the confluence of their actions. In other
words, if actions a and b are possible and do not conflict, then they are independent
and I can carry them out in any order, reaching the same end result; this is not the
case when there is conflict.
106 7 Consciousness

A deadlock with respect to a goal is a state where actions that allow for achieving
the goal are not possible. Thus, bankruptcy may be a fatal development for achieving
business goals, while the lack of freedom resulting from imprisonment limits the
possibility of achieving social and operational goals.
Free will is understood as the possibility of choosing between actions by applying
criteria resulting from the assessment of goals and the possibilities of
achieving them.
There are two extreme schools of philosophical thought in discord:
• One is “deterministic,” which asserts that our behavior is determined a priori by
various causes and internal choice mechanisms that we cannot comprehend. This
lack of understanding of causes is a source of an apparent uncertainty, which we
resolve when we make decisions. In other words, for given initial conditions, the
course followed by humans is predetermined. It is only human “stupidity” and
lack of knowledge that make people believe they possess free will.
• On the other hand, existentialist philosophers raise strong objections against
determinism. They (over)stress the ability of humans to make choices as one of
our defining traits.
Determinism views humans from the outside as machines and refuses to confer an
autonomous existence to mental phenomena, considering them a by-product of
external actions, the outcome of physical laws driving phenomena outside of and
within humans. Existentialism, on the other hand, views humans from within, acting
of their own initiative to a certain extent and affecting their environment.
In my opinion, what exactly free will is and whether it “really exists” is a
subjective ontological issue and is not amenable to rational analysis. Precisely
what our choices are, whether or not they are a self-delusion, is of no importance.
What is important is not just that our mind makes choices, analyses, and potentially
resolves conflicts between actions, but, more importantly, that we experience the
results of our choices.
Believing in free will is crucial with regard to our attitude toward life. We cannot
deny that free will is a phenomenon playing a crucial role in how we handle all our
relationships. People bear responsibility because they can choose. Machines bear no
responsibility. Their choices have been resolved by their programmers. Of course,
this does not stop certain foolish or ill-intended people from advocating for the rights
of robots and artificial intelligence!
How can the concept of human freedom be defined?
In theory, at least, when there are no needs, the issue of freedom is not relevant.
Many believe that freedom means to rid oneself of all needs or, at the very least, to
minimize one’s needs. This is why they stress that you have to even rid yourself of
the idea of need altogether.
Needs, of course, are inherent in humans. I believe that freedom from a need does
not mean that you are not subject to that need, but that satisfying the need is possible
within the value system in place. In this case, needs may vary in terms of their nature
and importance (physiological, safety, social, etc.).
7.3 Goal Management and Planning 107

A state of absolute absence of freedom is incompatible with human existence

itself. However, I may face a lack of freedom with respect to the ability to achieve
certain goals, for example, living a comfortable life, studying at university, ensuring
that I remain fed.
It is harder to judge when we can be considered to be free. One first answer is as
follows: when we have the choice of actions and meet their preconditions, and
therefore have the resources, to realize a hierarchically structured set of substantive
goals. However, caution is necessary here. With this definition, I must limit myself
only to optional goals. I must not include choices of actions with mandatory or
prohibitive goals that would be understood more as coercion than freedom.
The cost involved in achieving a goal must be commensurable to its importance.
However, even if I do not take the cost of achieving a goal into account that does not
mean that the more choices I have, the more freedom I enjoy. This is because, as we
have already discussed, the human mind is limited by cognitive complexity (see
Sect. 6.1.3) and cannot handle choices beyond a certain threshold. How this thresh-
old is defined depends on the abilities of each person; however, it is certain that
managing non-hierarchically structured choices beyond a certain number—let us say
five—proves ineffective.
Analysis of a large number of choices at a state results in loss of time and,
therefore, lack of effectiveness, if we wish to determine optimal solutions. An
approximate analysis might be preferable, being briefer and allowing for more
productive time to be used to carry out actions.
As is the case for computers, making choices also involves a computational cost
for humans. We must always strive to reach a compromise between the computa-
tional cost of our choices and the effective and timely execution of our actions. I
have seen people, mainly well-to-do, who have a plethora of options at their disposal
and, being unable to manage them, live a life of confusion and unhappiness.
In summary, freedom of choice entails a computational cost, which we must limit
for reasons of effectiveness. There are two crucial factors at play here.
1. One is the existence of a well-structured personal value system, which in a given
state, allows us to choose effectively among a large number of theoretical options.
It is no accident that the majority of ethical laws are proscriptive. There are people
who, in order to live “peacefully,” have imposed a large number of prohibitions
on themselves, from minor to major ones. For example, one might not drink, not
smoke, not drive, be a vegetarian, sleep every night at 10 pm and have a very
limited budget for food and clothes per day. In this case, the number of choices
definitely decreases. However, if a person feels comfortable living this way, they
may be happier than a person who enjoys more freedom but is unable to
effectively manage it.
2. The other factor concerns knowledge and its application in decision-making. Of
course, the knowledge learned at school, university or elsewhere, is important for
our ability to manage goals and prepare strategies. What ultimately counts,
though, is not knowledge in itself but the ability to manage acquired knowledge.
108 7 Consciousness

The above framework is generic and can serve as a basis for comparing different
situations. Even if a quantitative assessment does not make sense, there are clear
qualitative criteria that enable us to say that one situation is preferable over another.
Furthermore, the framework clearly indicates the role of ethics and ethical rules in
particular to simplify the freedom game, as well as the importance of knowledge and
“knowledge of knowledge.”
In conclusion, I would like to discuss the balance between individual and
collective freedom, which certain people often compare. I have noted that the social
and political environment limits the degrees of freedom a person enjoys.
There are two well-known extreme views: individualism and collectivism. The
former favors the unhampered action of the individual. The latter advocates for the
subordination of individual choices and actions to common goals. At this point, I
will not discuss what is obvious, i.e., that it is necessary to strive for balance between
individual and collective freedom.
Imposing numerous proscriptive/mandatory rules in a social organization creates
control mechanisms that hinder the performance and swift conclusion of social
processes. This imposition also limits the degrees of freedom for the action of
individuals, where they can exercise their initiative and inventiveness. Free spirits
find it difficult to express themselves and be creative in a totalitarian society. I have
witnessed how research productivity is limited in academic systems that are sclerotic
and “hyper-hierarchical.”
Mandatory/proscriptive rules must be minimal—provided, of course, citizens
know that they must make optimal use of their freedom. We must allow for optional
rules while also providing appropriate incentives for complying with them.
In the decision-making process, the present is a dividing line between the past and
the future, what has happened and what will happen. The present is the privileged
point where our conscious decision to act in a certain way or even to take no action
shapes the future. It is the point where what comes to pass takes shape, where the
past begets the future as the resultant of all actions, both conscious and subcon-
scious. It is the moment when information is channeled into the system and free will
is exercised.
Freedom is the conscious risking of the future and the existence of choices that
will shape it in order to achieve certain goals. The price of freedom is the complexity
of managing risks and choosing between conflicting actions. An absence of choices
means that the future has already been predetermined by the past.
7.4 The Conscious and the Subconscious Self 109

7.4 The Conscious and the Subconscious Self

7.4.1 General Considerations on Consciousness

7.4.1.1 Cooperation Between the Conscious and the Subconscious

I have discussed two systems of thinking, conscious and subconscious, and

explained that they work together in a hierarchical relationship: conscious thinking
is what normally overviews situations and issues orders (see Sect. 6.1.1).
Our perception of both systems of thinking is reflected, to a certain extent, in the
distinction we make between conscious and subconscious. The conscious looks
inwards; it “knows that it knows,” it “sees” possible choices of actions in the mental
system and makes decisions. The subconscious acts as an “auxiliary processor” to
fulfil goals achieved through automatic functions. At this point, I must underline that
I prefer the term “subconscious” to the term “unconscious” because, at least con-
ceptually, it better describes the way I understand the role of the “not conscious.”
I believe that our psychism was built gradually, like a game between our
conscious and subconscious. The question is how the conscious builds the semantic
model of the world that reflects the totality of our knowledge, and how conscious and
non-conscious functions work together in order to make decisions and realize them.
The mind consciously feeds and teaches automatic thinking to perform functions
that are essential for people, such as movement and speech. In fact, there are times
when the conscious cedes control to the subconscious and momentarily deactivates
itself in order to allow its unhindered autonomous functioning. When for example,
we play football or go dancing, we loosen our impulses and become “lost,” living an
experience that is intuitive and transcendental.
It has been experimentally proven that we need at least half a second to con-
sciously perceive and react to a stimulus. In contrast, the subconscious has a much
shorter reaction time. In certain experiments, where a weak stimulus is followed by a
strong stimulus over a period of less than half a second, the weak stimulus is grasped
by the subconscious, while it goes unnoticed by conscious thinking.
It is remarkable how certain people exploit the two-way relationship between the
conscious and the subconscious in order to influence our conscious judgment. I once
read that a psychologist tested his students by giving each of them an analysis of
their character based on their handwriting. Most of the students’ judgments were
positive, each of them saying how correct the analysis was. However, all students
had received exactly the same analysis—which proves the text contained the judg-
ments that each student would have liked to see, i.e., that they possessed robust
intellects, with a few weaknesses and feelings of insecurity, but that they knew how
to appear happy and strong.
Often, when faced with dangerous situations, the subconscious can take control
so as to allow us to react automatically and instinctively, temporarily blocking
rational, conscious analysis.
110 7 Consciousness

How do we strike a balance and self-control between the conscious and the
subconscious? For an athlete or artist, the subconscious is a key factor for their
performance. All transcendental experiences are characterized by this temporary loss
of the conscious, whether they are religious, erotic, artistic, or relating to creativity,
where the subconscious play an important role. An intervention by the conscious
could lead to confusion and hesitation and damage the quality of performance.
You might go to a concert and know the music by heart. You could listen to it by
playing a record at home. However, the result is not at all the same, as the physical
interpretation is transformed into a host of visible and auditory information that
carries you away. No emotion seems authentic when it is fully controlled by the
conscious.
We must reassess the role of consciousness in our lives. Many of the wonderful
things that people do go beyond conscious control. We can develop skills through
intuition and practice in order to communicate with the world and access cosmic
mystery. We can strike a fine balance between automated proscriptions and con-
scious choices. Automatic proscriptions help us overcome complexity, allowing the
conscious to focus on meaningful and creative options.

7.4.1.2 Consciousness and Computers

We know that conscious thinking has limited information-processing capacity and

accounts for only a small part of our behaviors and reactions. It is serial—we can
focus our attention on just one object at any time. I have also explained that the
complexity of relationships that the human mind can consciously grasp is limited.
Computers can obviously help us expand our intellectual capabilities. By their
very nature, they can analyze large amounts of data, playing a role similar to that of
the subconscious—that is, the role of “co-processor.” We can thus enrich our
knowledge and, indirectly, our mental capabilities.
The question is how human consciousness, given its inherent limitations, will be
able to understand and control the development of an increasingly complex and
rapidly changing environment (physical, technological). Unfortunately, humans are
currently unable to grasp the dynamics of scientific and technological evolution and
its impact. Systems have become so integrated and complex that it is no longer
possible to monitor their operations analytically and to assess their role and useful-
ness in detail.
Consciousness is the culmination of human evolution. Therefore, while it created
civilization that changed the world to a large extent, consciousness must not fall
victim to its own achievement. Consciousness that does not realize its limits, i.e., that
can only partially know the world, is a danger to itself. The civilization it created
affects its own evolution. We need to understand this fact and help the collective
conscience grasp the depth of changes and their consequences.
If you were to give technically sophisticated destructive weapons to primitive
people who have learned to resolve their differences using bows and spears, you
could cause a genocide—their collective conscience has not matured enough to
7.4 The Conscious and the Subconscious Self 111

measure the consequences of using such weapons in a heated conflict. During the
twentieth century, it seems that we avoided a global nuclear war by gradually
realizing that there would be no winners or losers in such a conflict. Averting an
environmental catastrophe and becoming more aware of the rational and controlled
use of computers are similar problems for our collective conscience.
We have confidence in big data and algorithms because they are supposedly
better than we are at making predictions. When faced with the question “What
should I study at university,” we may consider the answers of a computer more
seriously than our feelings and logic. When we are seeking a holiday destination or
choosing among candidates to appoint as members of a committee, we may consider
it more convenient to find answers using our computer. By making fewer and fewer
decisions in our lives, we are fostering a spirit of indolence and lack of self-
confidence. We feel relief by gradually shifting the weight of decisions onto
computers.
As it would be unthinkable to build without tools, it will be equally impossible to
manage the world without computers. However, the humanitarian vision that con-
siders humans a central value will de facto be tested.
What will the golden rule be? Not to allow computers to make critical decisions
by default. Not to allow, for the sake of “efficiency,” the creation of closed and
uncontrollable management and control circuits, of which only a few parameters can
be regulated by us. For technology to remain in the service of humans, we should
strive to rationally manage its complexity with increased awareness of our inherent
limitations.

7.4.2 Learning and Creation

Humans learn through automatic mechanisms and, at present, surpass computers in

this respect. The goal of education must be to foster not only critical thinking, but
also skills acquired empirically. Football players, pianists, researchers learn through
exercise and practice. Of course, during the learning process, they consciously apply
rational rules, but inventiveness in problem-solving depends on knowledge that is
difficult to analyze and theorize.
If a person can consciously solve a problem, they use acquired knowledge and,
ultimately, a comprehensible method that can be explained, transmitted to, and used
by others. If they solve the same problem without using reason and conscious
thinking, I have to assume that an equivalent process takes place in their mind. In
other words, there is an equivalence relationship, which at least holds true in logic
and the theory of computation.
At this point, we have to take account of Bergson’s views, which hold that direct
experience and intuition are more essential for understanding the world than science
and rationalism.
Indeed, while we can consciously analyze very little information (bandwidth), the
capability of the subconscious to absorb and consolidate information is much
112 7 Consciousness

greater. An image from a film, a painting, a line of poetry convey huge amounts of
information through possible interpretations—so much that we will never be able to
analyze them and understand what distinguishes a painting from a photograph. This
is where intuition comes in, which is the possibility of grasping and synthesizing the
“big picture” without resorting to analytical conscious thinking.
You can learn without profound theories, in practice, as they say. In order to learn
how to swim or ride a bicycle, you are not given a method to study and then attempt.
No matter how long someone explains it to you, you will not be able to swim in the
sea based on the explanation alone because you need to learn how to automatically
make certain moves without any analysis. When you learn how to play the piano,
certain logical rules are “internalized” and become automatic.
You cannot learn through theory how to become a good researcher—just as you
cannot learn how to become a good citizen. Of course, you can understand and
consciously apply certain rules; however, this does not suffice. There must be
assimilation into behavioral patterns. As the saying goes, “la culture, c’est ce qui
reste dans l’esprit quand on a tout oublié”—culture is what is left when you have
forgotten all you have learned—in other words, the result of thinking that has
ultimately become automatic, subconscious.
This raises the subject of controlled liberation of the subconscious, allowing it to
create masterpieces by subordinating it to a purpose. A great orchestra conductor, a
creator, impelled by passion and willpower, can reach soaring heights. That is where
the subconscious acts under the control of the conscious.
When you compose poetry, the creative process is largely subconscious, but the
criteria for what is beautiful and grand have been learned by reading poetry and
delving into the esthetics of language and meaning.
Artistic performance and creativity are intellectual achievements that require
perseverance and effort.
A few years ago, I read in a book how Avyssinos, a renowned Cretan fiddler,
whom I had the fortune to meet in person, learned to play the violin as a child. At
first, he travelled from his village to the big city for a week to study with a music
teacher. In the end, the teacher told him that, having shown him the basics, he had
but to do as follows: “You will lock yourself at home for forty days, you will not go
out at all and you will not do any work. You will just eat and sleep a little. You will
play the violin day and night, constantly bringing to mind the melodies you know.
As you play, try to carry them over to the violin. Once forty days have past, you will
be a fiddler.”
And it happened just as the teacher said. The young boy cloistered himself at
home for 40 days and when he emerged, he could properly play numerous
improvizations and local dance rhythms.
We may never come to understand exactly what the relationship between the
conscious and the subconscious is through analytical processes. It is just like our
relationship with computers. We interact, but cannot comprehend exactly how they
function—that is, how information is represented and transformed.
7.4 The Conscious and the Subconscious Self 113

7.4.3 Creation and Creators

I will conclude with a few personal thoughts on the process of knowledge creation.
How are new ideas born? How do we arrive at solutions to difficult problems?
There is a special moment during the creative process where the mind becomes
fertile and clear. Inspiration arrives like a spark, amazingly bringing solutions that
we might have been seeking through hours-long logical analysis. The emergence of
a solution possesses something magical. It is as if all our thoughts that might have
gathered in mind in disorderly fashion, suddenly organize themselves and show us
the way.
I recall what the great physicist James Clerk Maxwell said: “What is done by what
is called myself is, I feel, done by something greater than myself in me.”
It often happens that I wake up early in the morning with clear and creative ideas.
Of course, I do not believe that they dropped from the heaven and came to me “free
of charge.” They are the crystallization of thoughts that had been tormenting me for a
long while and now emerge in brilliant clarity. At times, I find it frightening; I
wonder how I can do it. Of course, the satisfaction of creation is ineffable and
wonderful.
It is also remarkable that, when I am considering a problem, my thoughts take
root deep inside of me and I cannot get rid of them—to tell the truth, I do not even
want to get rid of them—before I work on them in earnest and draw all the useful
conclusions. It is like a disease, a virus that settles in the mind and will complete its
“cycle” when I have fully explored the issue.
Creators are transformers of accumulated potential of experiences and knowl-
edge. They listen and interpret, decode, and translate the ineffable. They have built
their own world of symbols and representations and have their own paths that take
them far away to the sources of inspiration.
I cannot but recall Socrates’s “daimonion,” the internal “voice” that gave him
certain signs regarding his future behavior. A much more recent and characteristic
case is that of the autodidact Indian mathematical Srinivasa Ramanujan
(1887–1920), who died at the age of just 32 after making major discoveries in
mathematics. A deeply religious Hindu, he believed that he developed his results
based on ideas revealed to him by his family’s deity.
I admire the insightful intuition—for lack of any other term—of ancient Greek
philosophers. How did they make factual breakthroughs without experiments, with-
out even having visual or other technical means?
What amazes me even more is that the great truths are not the result of thoroughly
analytical or logically correct reasoning. The starting point is a driving idea (the
importance of numbers, the imperishability of ideas, the synthesis of opposites,
atomic theory, the continuous vs. discrete paradox, etc.) in support of which they
seek any manner of arguments. I remember, when I was a high-school student,
finding certain of Plato’s dialogs to be dull. They presented interesting ideas, but at
times, I had the impression that the thinking was incongruous or using flawed
114 7 Consciousness

arguments. What I mean to say is that Plato’s truly brilliant ideas stem from
“intuition” rather than being the outcome of a purely rational process.
A typical example is Plato’s conviction that planets, whose apparent trajectory in
the celestial sphere is complicated due to the movement of earth, follow a normal and
circular trajectory. The “argument,” if I remember correctly, was rather esthetic—he
could not fathom heavenly bodies behaving so irregularly! Another is the example of
the young uneducated slave, from whom Socrates elicits the solution to a simple
geometric problem in order to prove that the solution was already somewhere in the
boy’s mind, i.e., that learning is reminiscence.
How did ancient Greeks first lay claim to the freedom to think, to acquire
knowledge? How did the idea of theory emerge as a logical reduction of the complex
into a few predominant principles and elements?
In less than two centuries, we see a succession of intellectual giants, each of
whom would add to the foundations of the temple of knowledge. However much one
might disagree with the individual aspects, their conceptions are magnificent. Any-
one who over the centuries—and even in the present day—has tried to approach
major philosophical and scientific problems cannot but bring to mind these founding
fathers of philosophical and scientific thinking.
In conclusion, I would stress that inspired creation is a lonely endeavor. Although
creators may rely on the knowledge and help of a team, it is they who carve out the
path, who determine the framework, the vision. I am not referring just to artistic
creations, but also to designing an artifact, a major project which, even if it mobilizes
thousands of experts, always has one “chief architect,” a unit that inspires and
motivates all the actors.
The ability to create is, in my view, something innate that can certainly be
nurtured through education; however, knowledge alone is not enough. It is con-
quered through a process of deepening relationships between specialized items of
knowledge and linking them to a more general consideration of the problems of a
field of knowledge. It stimulates not only the conscious function of thinking, but also
the creative and subconscious.
Access to deep knowledge is a torturous, solitary endeavor. Irony and arrogance
are a veil before the eyes, putting us in danger of losing the way, leading to impasses.
However, like an experienced explorer on a dangerous mission, creators must be
vigilant so that they do not lose themselves in dead ends. I have seen talented
researchers trapped in dogmatism, prejudices, over-optimism, underestimating the
difficulties of penetrating knowledge.
Knowledge is conquered gradually. When you reach one peak, you see another,
much higher peak in the distance. This is a game without end. If you ever feel that
you have reached the end, you are done. You have lost the joy of ongoing creativity.
7.5 Seeking Happiness 115

7.5 Seeking Happiness

Much has been written about what happiness is and how to pursue it. The aim of the
discussion below is to say a few pedestrian and technical words relying on the mental
model presented above.
I understand happiness as an individual’s continuous capability to achieve his or
her goals by taking reasonable risks. This is a dynamic game of the mind with the
environment by developing appropriate strategies. It is dynamic in the sense that,
when certain goals are met, other new goals may arise. Therefore, the pace at which
goals are achieved must be commensurable to the rate at which new goals appear, so
that the player is not overwhelmed.
We can also imagine that certain goals are shifting. The closer we get to them and
partially achieve them, the more they become refined and enriched. For example,
building a house or achieving professional success are goals that become clearer as
we get near them and achieve them in part.
Happiness is to be moving toward reaching our goals by always remaining at
some optimal distance—neither too far nor too close. The optimum is determined by
the character of each individual. If I recall correctly, Kazantzakis said that each
man’s happiness is tailor-made to his measurements. There are those who do better
with far-off goals, while others are better suited to closer goals. However, the goals
must not be definitively reached because the game will then lose its appeal. I have
seen people who felt professionally fulfilled—and they grew unhappy or even ill
when they retired and suddenly withdrew from active living.
When you are too far away from where you would like to be, this becomes a
source of discouragement and frustration. Of course, certain people cannot set long-
term goals in their lives and are at the mercy of their immediate needs. This can
happen either because the resources available barely suffice for their day-to-day
survival or because they are incapable of mobilizing their intellectual and moral
powers consistently and continuously due to mental indolence.
Happiness is directly linked to the ability to understand the world and yourself,
but above all to create and envision. This is where education and background play a
crucial role. If the mind is feeble and remains untouched by the joy of all manner of
creation, then the game of happiness was lost before it ever began.
Although it may be adversely affected by subsistence needs and hardship, an
individual’s happiness is primarily a personal matter.
Collective happiness is not just the happiness of the individuals making up the
whole. It can be defined accordingly as success in achieving common goals that are
broadly and consciously accepted by the whole. Of course, collective happiness
affects individual happiness, and vice versa.
In conclusion, I would stress that I am conferring purely technical content on the
pursuit of happiness, and it has nothing to do with the moral values I will discuss in
the following chapter. Kant says that morality is not about becoming happy, but
rather about becoming worthy of happiness.
116 7 Consciousness

Happiness is a question of chemistry rather than logic. It is a balancing game

between what you are and what you want to be. We are not satisfied when we feel
inferior to what we would like to be. We say, “I don't want to be who I am – I feel
ashamed of myself.”
The same holds true in the opposite case, when we find ourselves at a loftier
position than where we can or dare to stand, when we have to play a role that requires
skills we do not possess. We sometimes say that someone falls short of a position
they hold and that this is reflected in their behavior, which is insecure and stressful.
Finally, there is also the tragic case of a lack of self-awareness, where our self and
our ego stand in ignorance of each other, i.e., we are not aware of our capabilities and
weaknesses.
I once wrote the following:
“Know thyself.” It is a prerequisite for harmony with ourselves in order to achieve
a happy life. It is the first and last major imperative for humans. All the others follow,
the way theorems follow from axioms. In our personal lives, knowing is not enough.
Unmanageable knowledge can become an unnecessary burden that not only does not
help, but also sometimes clouds our judgment.
Gaining knowledge through the game of freedom is the meaning of life. When the
cycle of life closes, our conscience must have discovered and fully understood the
rules.
Chapter 8
Value Systems and Society

8.1 About Human Societies

Human intelligence developed through various stages in which social organization

played a key role.
The words community and communication, both of Latin origin, are clearly
related. There can be no society without communication. The intelligence of living
beings evolved, thanks to social organization.
Societies are complex information systems.
Insect societies present impressive collaborative behaviors. They are character-
ized by a limited number of character types. For example, in bee society, these
comprise workers, drones, and queens. Each type has well-defined patterns of
behavior and communication, which determine their ways of interaction. Thus,
what characterizes insect societies is their collective behavior, with emphasis placed
on cooperation to achieve common goals of collecting food and reproducing. Insects
probably have no awareness of dilemmas and conflicts. They do not exhibit “selfish”
behavior.
In the case of mammals, we are moving onto a different evolutionary stage, where
animals have a strong feeling of self-preservation, which entails selfish behaviors
and “dog-eat-dog” conflicts. Such phenomena are intense in primates, where violent
conflicts are frequent and are limited only by fear of the stronger.
Human societies stand out from those of all other animals in that, through
evolution, they have developed organizational structures that allow for the coordi-
nated achievement of complex goals by defining the framework of actions of
individuals in the structures in question for their achievement. I have already
explained the importance of a common value scale and regulatory frameworks for
social organization.
We can regard human societies as complex information systems that possess all
the main characteristics of a mental system since they are theoretically the result of a
synthesis of individual mental systems. Of course, this synthesis takes place within

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 117
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_8
118 8 Value Systems and Society

structures and is regulated by the common value system. Thus, the analysis in Sects.
7.2 and 7.3 can be applied to a large extent to social groups, the difference being that
the individual’s consciousness is replaced by a more abstract concept: collective
consciousness.
Studying the dynamics of human societies is extremely complex. This complexity
is not just numerical, as is the case in physics when studying a large number of atoms
with exactly the same characteristics. It is “disorganized complexity” (see Sect. 4.
1.3), which depends on two factors.
On the one hand, each individual in society as a whole has particularities, which
cannot be summarized in a “type of individual,” as is the case, for example, with the
atoms of a gas.
On the other hand, individuals hold different positions in the social organization,
for example, due to income, occupation, or office, which affects their behavior. This
complexity makes it particularly difficult to study social phenomena, which exceeds
the aims of this book.
In this chapter, I will discuss how three important value scales take shape:
gnoseological, ethical, and economic, using concepts from previous chapters. I
will explain how the composition of each individual’s subjective experience takes
on an objective dimension, which can be studied as a social phenomenon. I will
discuss the institutions that play a fundamental role in shaping and maintaining the
scale of values. Finally, I will analyze the phenomenon of the deconstruction of
values and the decline this leads to, ending with a discussion of democracy in light of
the foregoing.

8.2 Gnoseological Values

Gnoseological values concern the validity of our knowledge about the world (epi-
stemic values), on the one hand, and the possibility of applying knowledge (techni-
cal values), on the other. These are common social values reflected through the filter
of subjectivity on individual value scales. Consequently, gnoseological values can
be judged according to objective criteria, as well as in relation to their degree of
acceptance by society as a whole, which I will discuss in what follows.

8.2.1 Epistemic Values—Truth and Falsehood

A key property of knowledge is that it is “true”—otherwise, it would lose its

utilitarian value—i.e., it helps us understand and change the world (see definition
of knowledge in Sect. 3.4.1).
Truth is an epistemic value and characterizes the concordance with reality and the
rules of reasoning. Falsehood is therefore defined as the opposite of truth. Here we
must make it clear that we usually confuse falsehoods, being “untruths,” with lies,
8.2 Gnoseological Values 119

which are a conscious distortion of the truth. An untruth characterizes information

that is not empirically or logically substantiated, while lying is an ethical value.
A person knowingly telling a falsehood is lying and thus has a moral responsi-
bility. The phrase “the earth is flat” is false. If someone in the Middle Ages were to
claim the phrase was true, then they were possibly not lying, because it reflected their
level of knowledge. In contrast, if a present-day educated person were to assert that
the phrase is true, then they would be lying.
Nevertheless, we cannot certainly ascribe a value of “true” or “false” to every
sentence that constitutes information regarding the world. This is generally due to
our inability to precisely define the meaning of natural languages. When we talk
about social or economic phenomena or express judgments about people’s behavior,
our propositions may be open to multiple interpretations and it is therefore impos-
sible to ascribe the value of “true” or “false” to them.
However, even in the field of mathematics, where language is strict, there are
renowned undecidable propositions, which we discussed when referring to Gödel’s
theorems (see Sect. 3.4.2). There are also inherent limitations to knowing the truth
about physical phenomena, as we have explained in detail.
The truth has both an objective and a social dimension. Its emergence takes place
through a complex social process. What a given society considers to be “true”
reflects the level of knowledge at the time, shaped through filters controlled by
social institutions. Perhaps someone would say that I am kicking at open doors by
stating the obvious. Everyone knows the story of Galileo and the other “heretics”
who came into conflict with established views. In this case, the emergent truth
challenged the authority of the Church.
Every kind of truth, in order to be adopted by the many and become a social
value, must be subjected to a process of certification, where the social institutions
play a crucial role.
Let me use a completely “neutral” example as an explanation. In the mid-1990s,
there was intense debate around the proof of French mathematician Fermat’s last
theorem. This is an extremely difficult problem that remained unsolved for
358 years, despite many asserting from time to time that they had solved it. In
1995, the British mathematician Andrew Wiles published a correct proof. How is it
that we now believe that this problem has been solved? This is because groups of
mathematicians—whether or not organized, it matters little—have read the long and
highly technical proof that Wiles published and agreed that it is correct.
Similar processes occur with all kinds of emergent knowledge. In order for the
theory of relativity to become common knowledge and established, it took debates at
academies, publications in journals, books, and newspapers. This is how it became
knowledge as part of the educational system and entered the public domain.
Unfortunately, in the knowledge and information society, institutions such as the
press and academic organizations do not play the role they should be playing as
custodians of truth, nor do they help properly inform the public of all the major
problems of concern. This poor and sometimes distorted provision of information
poses a huge risk to individual and social liberties.
120 8 Value Systems and Society

Increasingly, fabricated lies become a tool to mislead and manipulate the many.
The powerful falsify or hide the truth, spread false rumors, use rhetorical figures to
impress and mislead. The internet and social networking are multiplying their
devastating effects.
This is particularly the case when people, being listless and disoriented, do not
care to protect themselves. In fact, people have the natural tendency to believe
anything unbelievable and absurd that stirs up emotions and passions, which culti-
vates illusions. They close their ears to the truth and real problems that disturb and
upset them. At present, the concept of truth is at risk due to confusion and familiar-
ization with falsehoods, due to the dulled judgment of the many.
Finally, I would like to say a few words about what has been called “post-truth.”
The concept describes the lack of common criteria as to what is true and what is not.
It states that fiction counts more than facts. Why is this?
I think that what we are experiencing is the result of two factors in particular. On
the one hand, it is the moral crisis and relativism, which blur the boundaries between
truth and falsehoods. On the other hand, it is the fact that public opinion is being
bombarded by a wealth of information of all kinds, and that it is difficult to
discriminate between unsubstantiated “news” and corroborated facts.
Public opinion tends to uncritically believe rumors and fabricated news. People
are fed up with and saturated by the “real stories” served up by all types of media on
a daily basis. Promises that are not kept, manipulations and deceit lead to “irrational”
behavior.
How can we explain Brexit or the Capitol attack in January 2021? How can we
explain the boom in religious fundamentalism and religious conflicts that have been
testing Western societies since the late twentieth century? How can we explain the
powerful anti-vaccination movement being observed in some European countries,
placing the lives and liberties of thousands of people at risk on unfounded and fear-
based arguments?
Societies are simply tired of the parody of modern democracies and their inability
to solve blatant problems. It is a failure of the elites to shape and realize a vision for
stability and prosperity.
How long will this situation last, and how will it evolve? It is difficult to predict
whether this is a temporary absence of rational thought or a permanent triumph of the
absurd.

8.2.2 Technical Values

8.2.2.1 Correctness and Erroneousness

While the truth involves concordance with facts and logic, “acting correctly” means
taking a step toward achieving a goal, which is defined by specific needs. An error,
therefore, is any deviation from a goal. It is integrally linked to our mind’s inability
to properly assess a situation, to aim right.
8.2 Gnoseological Values 121

“Correct” and “erroneous” are technical values. Error etymologically originates

from the Latin verb “errare,” which means to roam, i.e., to stray from the right path.
Technical values help us assess whether an action is successful according to
gnoseological criteria. They also allow us to weigh the responsibility of the actor:
this is the case when an error on their part causes damage and they are found to be
negligent because they did not properly apply their technical knowledge. Thus, any
expert, whether an engineer, a doctor, or a cook, may make mistakes when applying
their knowledge. These result in professional liability, for example, when a building
designed by an engineer collapses or a surgeon endangers the life of a patient due to a
manifestly incorrect action or omission.
While what is true is assessed according to objective criteria, what is correct is
judged in relation to goals and their feasibility. These goals must be clearly formu-
lated and, of course, must not contradict the legal and ethical rules in force. In any
case, knowledge of the truth is a prerequisite for us to act correctly.
The most assured way to make no errors is to do nothing. Both the prudent and
the foolish make mistakes. However, prudent persons have the foresight to anticipate
the impact of their errors and to take measures to prevent critical situations. They
may also deliberately make minor mistakes in order to investigate a situation and
learn—“paying to learn,” as is the case in poker, in order to understand the
opponent’s tactics. However, prudent people make sure that their errors are not
irreversible.
As important as the truth is, so is social highlighting and certification of what is
correct. That is why there are institutions that play a controlling and validating role.
We board airplanes, live in skyscrapers, use electrical appliances with the conviction
that they are quite safe because they were designed based on theories and rules;
because certain independent authorities have certified that they have been
manufactured according to correct procedures and specifications.
I have pointed out that, while standardized certification procedures are applied for
each product, such as children’s toys or electrical tools, there are no such procedures
for computer systems and their services. No certification is required to ensure that
they function properly with respect to specific safety and security rules, except for
certain critical applications, such as transport, nuclear power plants, and smart cards.
This situation engenders the risks already discussed in Sect. 6.4.2.

8.2.2.2 Risk Management Principles and Their Implementation

I explained that managing goals is linked to the concept of risk, which characterizes
the probability of finding ourselves in states that range from hazardous to cata-
strophic, with high value costs (see Sect. 7.3.1).
Let us use the coronavirus pandemic as an example to discuss risk management
principles. The different way each country has handled the pandemic highlights the
complexity of the problem of assessing risk and reaching compromises between
safety and liberty (see Sect. 7.3.2).
122 8 Value Systems and Society

There are two ways of approaching risk management.

1. An approach is based on the precautionary principle, which “requires a rapid
response to be given in the face of a possible danger to human, animal or plant
health, or to protect the environment,” as described in European legal texts
[1]. This approach is appropriate in critical situations. Where the scientific data
do not allow for a full assessment of the risk, radical measures must be taken on
the basis of the “worst-case” scenario. The principle is applied in cases of
catastrophic weather phenomena, epidemics, and other natural disasters. It is
also applied in order to assess the risk of using pharmaceuticals, medical devices,
and means of transport, the development of which is subject to control by
independent organizations. The precautionary principle is based on another,
more fundamental principle governing the European legal framework: that life
is of the utmost value. When a human life is at risk, the cost required to save it is
unimportant.
2. The second approach to risk management is applied when the criticality of what is
at stake is minor or, at the very least, does not concern human lives or severe
environmental disasters. It concerns the optimization of criteria, where
conflicting factors that characterize an activity arise. If, for example, the risk
(i.e., the probability of something harmful occurring) increases in proportion to
the intensity of an activity, then I can determine, if I have the appropriate
mathematical model, the optimal level of activity. Therefore, if I am driving on
a motorway, there is, theoretically speaking, an optimal speed to reach my
destination which is a compromise between travel time and risk. The optimum
depends on many objective factors such as weather conditions, traffic, and car
type, as well as subjective factors, such as driving ability. The optimum corre-
sponds to the level where the benefits are maximized in relation to the associated
costs due to risk. Finding optimal operating points for business risk management
is one of the fundamental problems of economics.
These two ways of addressing risks are reflected in the management of the
coronavirus pandemic. At first, the leaders of many countries behaved more or less
as managers, taking the second approach. Human life was simply the adaptation
parameter of an equation, where stock market, economic and political criteria were
taken into account.
The application of the precautionary principle would require the immediate
imposition of strict measures aimed limiting the spread of the virus so as to save
as many lives as possible. However, this was not the case except in very few
countries. Thus, while the contamination of the population in the United Kingdom
was out of control, the British Prime Minister was talking about “herd immunity.”
We understand that he prioritized keeping economic activity at a satisfactory level at
any cost. He did not take into account the cost in human lives when 60% of the
population would have been infected. Therefore, even if a small proportion of
patients lose their lives that translates into tens of thousands of lives.
A similar attitude was shown by the USA, where President Trump announced that
the threat to “the population and economy” was minimal and referred to a “Chinese”
virus. It is interesting to note that fear of the pandemic’s impact on the economy was
8.3 Ethical Values 123

at least as significant as fear of loss of human life. The then Governor of New York
State, Andrew Cuomo, made statements against Trump on 28 March 2020, claiming
he did not have the right to put the city in quarantine, characteristically stating that
“New York is the financial sector. You geographically restrict a state, you would
paralyze the financial sector.”
However, even in other countries closer to us, governments were unacceptably
flippant, adopting a “wait and see” attitude. France held the first round of municipal
elections on Sunday, 15 April 2020, and was forced to immediately cancel the
second round the following day.
Given the foregoing, I would also like to highlight a dangerous change in attitudes
to crisis management, which was gradually adopted by the centers of power over the
course of the last few decades. This marks a transition from strict adherence to
precautionary principles in the 1970s to adaptive management, where human life is
not what is most important, but just a parameter in a game dominated by economic
and political criteria. Unfortunately, we moved from a time when at least the major
Western democracies were quite robust and had the means to implement their policy,
to an era when even the policies of the most powerful state are tied to the whims of
the markets.
Similarly, with regard to system security issues, I have had the opportunity to see
this slackening and transition from the precautionary principle to a financial-style
risk management, even when human lives are at stake.
I have explained that the USA currently accepts self-certification for critical
systems. This means that there is no safety limit certified by an independent and
responsible authority, but that this limit is determined by the manufacturer itself. One
characteristic case is that of the Boeing 737-Max, which had been self-certified with
a single angle-of-attack sensor, while regulations require at least two sensors [2].
This policy change was considered necessary because the precautionary principle
makes the widespread use of new technologies, such as artificial intelligence and
autonomous systems, economically prohibitive.
The way the coronavirus crisis is being managed is indicative of the revival of a
systemic ideology of barbarism at the global level. It highlights the weaknesses of a
system, which are increasingly visible in crisis management, such as climate change,
but also control of the use of new technologies. The exclusive priority of the
economic and technological factor irreparably jeopardizes the foundations of our
civilization, the cornerstone of which is respect for human beings and their lives. I
am afraid that the price of barbarism will be even heavier in the future.

8.3 Ethical Values

Ethics is the area of philosophy that studies the concepts of good and evil.
Ethical values are crucial for social organization and are therefore assessed on the
common value scale; at the same time, they are also individual values. They would
be completely useless in societies without friction and without uncertainties. They
124 8 Value Systems and Society

contribute to conflict resolution and achieving consensus, mutual trust, and predict-
ability, which are essential for social peace and progress.
Complying with ethical rules is a matter of choice and therefore depends on each
individual’s sense of responsibility. Sometimes this stems from the need for recog-
nition by society as a whole when society values it and rewards such compliance.
Throughout history, modern societies have adopted a relatively comprehensive
set of ethical rules and values.
Ethical values emerged alongside the first human societies. Initially, their devel-
opment was inherent in the development of religious sentiment. Later, ethics became
the subject of rational study by philosophers, starting with Aristotle.
Contrary to gnoseological values, ethical values cannot be based on objective
criteria. Their adoption is the crystallization of best social practices or is a question of
faith (belief)—but this does not mean that ethics lacks rational foundations.
In Sects. 8.3.1 and 8.3.2 below, I will discuss the concepts of good and evil and
the role of faith in laying the foundations of ethical rules and managing ethical
choices.

8.3.1 Good and Evil

Good and evil are defined as right and wrong, respectively, in terms of ethical values
and rules. Managing good and evil is the central ethical problem facing our con-
science. Any breach of ethical rules is considered to be bad, for example,
non-compliance with mandatory rules or violation of proscriptive rules.
Usually, there is some distance separating good and evil. They are concepts that,
while defined by subjective criteria, also have a broad objective basis in the common
value system.
It is no accident that the golden rule of ethics is common across many peoples:
Cleobulus of Lindos wrote “ὃ σὺ μισεῖ ς ἑτερῳ μὴ πoιήσῃς,” while the Old Testa-
ment includes the phrase “ὃ μισεῖ ς, μηδενì πoιήσεις”—both essentially meaning “do
not do unto others what you do not want them to do unto you.”. This is an admirable
empirical rule of balance that ensures harmonious coexistence in societies.
Many ethical rules have a deterrent or prohibitive character, such as the rules of
the Mosaic Law, “thou shalt not murder,” “thou shalt not commit adultery”, and
“thou shalt not bear false witness.” These rules are also of practical interest: they
reduce individual choices and complexity in managing freedom.
In theory, ethical rules seek to ensure that people co-exist harmoniously and work
together in a given social context. As such, they are not independent of religious and
other beliefs. Furthermore, they tend to shore up the status quo and the rights of the
powerful.
In the Middle Ages, “buy cheap and sell high” was a grave sin, constituting moral
misconduct and, at times, even a criminal offense. Seeking riches for the sake of
riches was reprehensible, and flaunting one’s wealth drew the ire and condemnation
of society.
8.3 Ethical Values 125

Today, seeking profit by all means is regarded as evidence of gumption and, to a

certain extent, a virtue. The nouveaux riches of recent decades in particular, not only
do not hide their wealth, but also appear proud of their power and their vast fortunes.
Certain ethical rules aim to establish relationships of trust that allow for the
predictability of behavior. This is absolutely essential for the convergence of the
actions of individuals in the pursuit of common goals and the effectiveness of
transactions. Such rules include not lying, keeping one’s word, not being
hypocritical, etc.
Other rules require that our actions be inspired by a sense of fairness, i.e., not to
put our personal interest above all else, ignoring the impact on others. This means
courtesy, respecting the choices of others, not unnecessarily excluding them and
disregarding them.
When we are selfish, even when pursuing a small personal interest, we can harm
others by hindering the achievement of common goals. In contrast, when we are
being altruistic, when we overcome our personal interests and, in certain cases, when
we make personal sacrifices, we contribute to others and society at large.
Other ethical values concern how we choose actions, such as a sense of respon-
sibility and courage. You are responsible when you choose actions that not only do
not violate ethical rules, but also lead to the best result. There are many who do no
harm, but few who, out of loyalty to and belief in the common good, try to do what is
best at every turn.
You are being courageous when you have confidence in your strength, self-
discipline, optimism, and perseverance in achieving goals.
Evil violates ethical principles, while anything that does not run contrary to
fundamental ethical principles and contributes toward individual and collective
happiness is considered to be good. Of course, both depend on the circumstances
and have range and intensity. They are judged on results in the short and long term.
A small good can result in a big evil over time and vice versa. One major difference
is that good generally takes effort, while evil can also be done when one does
nothing. Neglect and inertia can cost dearly, as time waits for no one.
A virtuous act is one that results in positive ethical values as defined by the
common scale of values. Conversely, the absence of virtue characterizes actions that
result in negative values. Thus, we speak of cowards, egoists, criminals, etc. As I
said, selfishness is manifested through actions that, above all else, lay down criteria
for meeting personal needs, ignoring the effects of these actions on others. Acting
illegally may be ethically reprehensible but is characterized as non-compliance with
the law.
The absence of virtue at the ethical level leads to a reduction in self-esteem, which
can manifest itself in feelings of shame or guilt in certain individuals. Shame is the
feeling that we did not rise to the occasion and violated ethical rules. Guilt concerns
acts that are usually considered criminal by a legal system. Guilt may result in
remorse, a feeling deriving from fear of the punishability of the actions in question.
Often the pursuit of good can bring us into conflict with others, especially those
who pursue evil. This is the case with the heroes of antiquity, the prophets of the Old
Testament, and the enlightened individuals we celebrate as martyrs. Hercules fought
126 8 Value Systems and Society

evil during his 12 labors, just as Theseus punished the brigands and freed the
Athenian youths from subjection to the Minotaur. Martyrs sacrifice themselves,
actively resisting evil by fighting for human and national ideals or by not accepting
to yield and deny the ethical values dictated by their faith.
It is our moral obligation as human beings, not only to do good, but also not to
give in to evil.
I must point out here that there are certain gray areas. A “good evil” idea is firstly
an evil one, i.e., it has harmful consequences or engenders risks, but in some
circumstances, it becomes so attractive that its glamor eliminates the inhibitions
regarding its execution. Examples include stealing from the rich to give the poor,
breaking things during a demonstration in order to claim what is right. I believe that
ethical goals must only be achieved by ethical means. The opposite approach can
lead to dead ends and vicious circles.
“Evil good” ideas are more dangerous. They are popular, enchanting ideas whose
glamor can seduce us into hurriedly and uncritically realizing them, without thinking
about what their realization might cost or whether they are even feasible.
These ideas set goals that, in theory, meet high ethical criteria, such as absolute
social justice, and whose implementation leads to disasters or dead ends. These are
“blanket” ideas, which in the name of misconceived equality, limit creativity, and
individual initiative and do not accept the diversity of individuals. They feed
populism and demagogy.
We can find numerous examples of “evil good” ideas in all aspects of life,
especially in political and social life, such as educational reforms everyone is in
favor of, debt forgiveness, and early retirement at the age of 50. However, this is also
the case even in scientific research, where vast amounts are spent uncritically on
ostensibly lofty but manifestly infeasible goals, or where the expected benefit is not
commensurate with the cost entailed.
The relationship between the conscious and subconscious in the sphere of ethics
is also interesting. Once a good Israeli friend, who unfortunately left this plane of
existence early, made the following interesting statement: the actions of a good
Jew need be in line with the Laws and Commandments; how he thinks is indiffer-
ent in God’s eyes. There is a difference here with the Gospel message, which
condemns both the sinful act and the thought: “And if thy right eye offend thee,
pluck it out, and cast it from thee,” “And if thy right hand offend thee, cut it off, and
cast it from thee.”
This difference had long since given me pause, and I believe I have an answer
regarding its importance. For a Christian, doing good must be a natural disposition
and not an obligation, it must be something automatic and spontaneous. This is
because their conscience has nourished their soul and the Christian does good as
easily as breathing, as naturally as their heart continues beating. You do good not just
to show God that you are following His law. Doing good becomes a need, not a
coercion. It is inherent to your existence.
US President Jimmy Carter once fell into a trap, during an interview on the
campaign trail, when asked whether he had ever cheated on his wife. He answered:
8.3 Ethical Values 127

“I’ve looked on a lot of women with lust. I’ve committed adultery in my heart many
times.” This stirred up a storm of disapproval and protest!
There are strong desires that have been pushed down to the subconscious and no
longer subject to willpower, for example, when you decide to quit smoking perma-
nently or decide to become a vegetarian. These automatic rejection mechanisms,
which we create at a time through our willpower, simplify the game of choices. This
can make our life easier or more pedestrian, depending on the situation.

8.3.2 Faith and Dogmatism

“How can a scientist be a believer?”, someone once asked me. He happened to be a

mathematician, so I reversed the question, asking him if he understands why axioms
are needed in mathematics. I said that my beliefs concern questions that are beyond
the realm of knowledge. Believing does not mean that I do not accept the scientific
approach. Belief complements knowledge and is necessary for achieving coherent
behavior. He insisted and was absolutely convinced that any kind of faith is a
prejudice from which all rigorous thought is naturally free.
Faith is a concept that has been completely degraded in these times. This may
well be because it became worn out over the centuries, due to our having loaded it
with much more than it could bear. Then again, perhaps it is because we are living in
a time when opportunism is considered a virtue.
A twentieth-century French politician, who stood out because he occasionally
served various political ideologies, said in jest “it is not the weathercock which turns;
it is the wind!”
If we consider the a priori acceptance of certain beliefs as faith, then no approach
to knowledge is free of this consideration. Mathematics is based on axioms. The
development of any scientific theory is based on hypotheses that, of course, must not
be contradicted by reality.
Faith in ethical and spiritual values can be considered a set of beliefs, a type of
“axioms” that cannot be directly validated. However, we are able to judge them by
the result of their implementation. Even religious beliefs and the resulting practices
are tested through personal experiences. Each individual possesses a metaphysics,
whether they admit it or not.
As I have explained, the mind uses the individual scale of values and relevant
rules as knowledge for assessing needs and making decisions. You cannot build a
logical construction without points of support. Without such support, the mind
flounders and gets nowhere.
Therefore, adherence to ethical principles also stems from a practical need to
effectively manage one’s liberty. We can confer a purely technical character upon
this adherence.
What we call dogmatism is something different altogether and has nothing to do
with adherence to values. Dogmatism means giving your convictions a basis, which
is unquestionable and cannot be debated, and trying to impose these convictions at
128 8 Value Systems and Society

times. We must fully respect the convictions of others, insofar as they do not
establish any infringing practices or offend public sentiment. Therefore, freedom
of religion must be a right in every society founded on rule of law.
In conclusion, I would say that it is better to have a wrong faith than a lack of
points of reference and an ambivalence. An intelligent, honest person will discover
in practice that adherence to the wrong convictions creates practical problems and
will revise them accordingly. We must not look at things in a static manner.
Adherence to ethical and spiritual values and faith in the common good is not the
result of divine inspiration. It is not something that someone places in your head and
stays there once and for all. Convictions are created or lost, are jeopardized and
challenged every day in the dynamic collective game involving people’s actions,
behaviors, and relationships—political, economic, social, from the simplest to the
most formal.

8.4 Economic Values

Distinct from that of all other species, human social organization is based on the
cooperation-conflict model. Economics considers human behavior as a relationship
between given needs and limited resources with alternative uses. Needs and
resources are matched through market mechanisms, which dynamically determine
economic values by means of the interaction of supply and demand affected by
consumer behavior trends and the role of institutions.
The study of the economy, as we know it, dates back to the origins of capitalism.
Before economic laws could be understood and implemented, many changes had to
take place.
At first, economic relations were defined by traditions and religious prejudices,
which aimed at maintaining harmonious coexistence in societies. For example, the
caste system in India aimed to strike a balance between professions. A similar
stratification system was employed in ancient Egypt and in ancient China. The
idea of profit for profit’s sake was deplorable in the Middle Ages. For centuries,
guilds played a major role in urban life and exercised a stranglehold on market
economies by protecting their corporate interests.
Money is one of the most powerful abstractions of the human intellect. Every unit
you acquire is a promise that you will have goods and services, that others will offer
so that you can purchase anything for sale in the future. You can therefore purchase
confidence for your future consumption and freedom of time, but you cannot
purchase a future with your money, that is, the actual time to live your life. This
is—fortunately—a fundamental limitation of money. It would be a nightmarish
world if one could buy life, for example, take another’s life counter down to zero
and add it to one’s own.
What I want to stress is that the economic game has an important informational
dimension, where ethical and technical values are crucial. The effectiveness of social
organization is equally crucial. It depends on mechanisms for establishing ethical
8.4 Economic Values 129

values, the legal and regulatory framework for transactions, the processes of wealth
creation and redistribution, and the management and development of knowledge.
Transactions require mutual trust to reduce risk between the parties involved and
to enable seamless and secure communication, which is currently achieved through
authorized bodies. This view is confirmed through the rapid growth in transactions
carried out via computers and telecommunications. It negates the one-dimensional
standpoint that only emphasizes the strictly material nature of exchanges.
The first banking systems were developed in the West during the Crusades and
subsequently through the Jewish diaspora. The existence of institutions to facilitate
and guarantee transactions between third parties and thus profit is crucial for the
development of commerce and finance. Central banks, through their dual mandate as
independent authorities, shape monetary policy and oversee banking institutions.
In the future, the use of blockchain technologies will allow for transactions
between parties without third-party guarantors. This may threaten the existence of
banking institutions causing them to surrender centralized control unless they adopt
blockchain technology, particularly in order to address complex cost and operational
challenges.
However, the lack of central arbitration and guarantee bodies entails high com-
putational costs. It also makes transactions more vulnerable to cyber-attacks. That is
why it is difficult to predict what the future of these technologies will be. If they
ultimately replace central banking and control institutions, this will not just seriously
impact the organization of transactions at the international level but will also be an
important step toward deeper globalization.
I will end with a few thoughts on the evolution of the dominant economic model
toward what we have been calling neo-liberalism since the mid-1980s. This call for
“less state” in the productive sector brought about a series of reforms in the Western
world accompanied by massive privatization, with the state being reduced to playing
a purely regulatory role so as to allow the “natural,” smooth functioning of the
market, at least during growth periods.
Most states no longer have sufficient financial resources of their own and drivers
necessary to intervene in a forward-thinking manner. They do not have a “nest egg”
for difficult times and must raise funds by borrowing from international markets. In
my view, this premeditated weakening of states will serve as yet another step toward
broadened globalization and the disappearance of the “nation state.” Countries such
as China and Russia, which are protected by a kind of state capitalism, may escape
this trend.
At the same time, we are experiencing the internationalization and integration of
commercial and financial relations, thanks to the opening up of markets and the
computer revolution. This led to the rapid growth of big tech companies such as
Google, Apple, Facebook, Amazon, and Microsoft, which boast enormous eco-
nomic power and influence. The global dimension of these businesses has allowed
them to escape the control of states in terms of paying taxes and complying with
regulations and laws.
Neo-liberalism goes a step beyond slackening social ties and the rejection of any
idea of interventionism and programming. It strives for competition, the creation of
130 8 Value Systems and Society

inequalities, and the reduction in social services. All this as a means for creating
markets, which entails destroying collective structures and fostering a culture of
individualism.
This leads to a deification of the market and its automated processes.
Humanity has suffered from the economistic view that all social phenomena can
be analyzed based on the economy and its dynamics, focusing on the economic game
and disregarding the contrasts between the individual and collectivity, humans and
nature, nations and races. Unfortunately, this spurious view serves as a meeting point
for both historical materialism and neo-liberalism.
Neo-liberalism proposes a kind of “cybernetic” vision of a market-driven world:
efficiency is achieved through the automatic interaction of economic “machines”
that exchange information and are motivated by profit. Von Hayek, one of the
instigators of neo-liberalism, says that the market is an information processor more
powerful than the human brain. I quote: “the market is posited to be an information
processor more powerful than any human brain, but essentially patterned upon brain/
computation metaphors” [3].
I cannot know where such worship of the market can lead humanity. However, I
am not optimistic about the outcome.

8.5 The Role of Institutions

Since their inception, societies have been based on the existence of institutions.
These are structures that contribute to the imposition of social order by promoting
the common value system and the application of explicit or implicit rules. Explicit
rules are laws, regulations, and institutional frameworks that prescribe the actions of
individuals and bodies. Implicit rules characterize the attitudes and morals of the
social environment. They play a major role in the successful implementation of the
explicit rules. One characteristic example is the attempt to apply the US constitution
in Latin American countries, which failed miserably due to the different social
environment.
Through their actions, institutions define, to a large extent, the scale of values of a
society: what is true and what is not credible, what is right and what is wrong, what is
legal and what is illegal, what is good and what is bad. They codify balances, which
take account of the social and political conditions that have developed over the
course of history. Their proper functioning and adaptation to shifting situations are
essential for the survival of a social system.
For each field of action—economic, political, legal, educational, military,
gnoseological, ethical, religious, esthetic, etc.—there are corresponding institutions
that define and promote specific purposes and values, which all work together to
achieve social order and some idea of the common good.
In primitive and ancient societies, interactions in relatively small social groups
were personal. Misconduct brought shame to the perpetrator and their family. Such
relations, which were still essential in democracies such as ancient Athens, were
8.6 Depreciation of Values and Decline 131

definitively lost upon the disappearance of the city-state and the creation of empires,
such as those of Alexander the Great or the Roman Empire.
In modern societies, the state has a responsibility to renew and adapt existing
institutions, and to develop a vision that will safeguard the future of society as a
whole by planning actions for growth and prosperity.
Institutions form public opinion to a large extent and foster social cohesion and
synergy. What is the state of the world economy? How does a country’s education
system perform? Is artificial intelligence dangerous? How are we dealing with
climate change? Is a vaccine dangerous and, if so, how much?
History is full of examples where misinformed or insufficient public awareness is
swayed into making wrong decisions with devastating results.
We see that the current crisis of values reflects and is reflected in the inability of
the institutions of national and international organizations to rise to the
circumstances.
One blatant example is the failure of the United Nations and, more recently, the
World Health Organization, to respond effectively to crises with a view toward the
common good. At the same time, the rapid development of science and technology
has made some believe that we will live in an increasingly better and more control-
lable world. However, progress alone cannot guarantee that human societies will be
able to adapt and effectively address the challenges emerging, which are increasing
automation, overpopulation, pandemics, and climate change. International coopera-
tion and the existence of effective international institutions are more necessary
than ever.

8.6 Depreciation of Values and Decline

How does a society “decide” to die? How are nations, civilizations, and empires lost?
How does a society break down into, but a patchwork of people impelled by their
tendencies and appetites with no internal cohesion, easy prey for malicious and
predatory people?
There are numerous examples of societies that declined and collapsed not so
much due to external risks or economic crises, but because they ceased to function as
coherent social groupings. One characteristic example is the decline of the Roman
Empire, which was due to a crisis in ethical values, and the progressive degradation
of the value system on which the res publica was based.
Societies function well as long as there is strong cohesion around a consistent,
common scale of values, as long as individuals consent to sacrifice part of their
personal liberty in order to achieve common goals.
Every social crisis is the result of the depreciation of the common value system in
its various fields. It is not easy to understand how fields affect each other. Depreci-
ation is the result of a general slackening, a mismatch between established values
and newly emerging ones. Of course,—a rare occurrence—the crisis can prove
132 8 Value Systems and Society

fruitful when the emerging values form the basis of a powerful vision for change,
such as the French Revolution.
It is interesting to examine how certain countries that had all the qualifications for
growth and prosperity, declined and suffered a profound economic and political
crisis. One characteristic example is the case of Argentina, a country vastly rich in
natural resources that experienced an exceptional period of prosperity and growth
comparable to that of advanced countries in the first half of the twentieth century.
Since then, it has fallen into a cycle of successive crises and decline.
More or less similar situations show that factors relating to social and political
organization, the functioning of institutions and the cultural level of people are more
important than material goods. Many underdeveloped countries have the material
conditions for progress and prosperity. Economic determinism, which places pre-
dominant emphasis on the economic factor, has failed miserably.
Even more important than economic resources is the ability of a society to act as a
coherent (information) system capable of effectively bringing its creative powers
together around realistic goals and planning to achieve them.

8.6.1 Decay of Language—Degradation of Concepts

Values start being undermined when language and concepts are degraded. If a
language grows poor and concepts decay, we lose the corresponding possibilities
of formulating knowledge.
Language is one of the most valuable common goods. Its erosion opens the path
to decadence and decline. When words lose their common conceptual value, they
become like a useless, counterfeit, devalued currency. Society’s capacity to function
as a whole is degraded, and its cohesion is eroded by the cultivation of an ideological
climate, which opposes rigor and seeks ambiguity. The rules and criteria for rigorous
thinking and clarity are lost. As clarity of speech fosters a climate of trust and honest
trading, ambiguity opens the door to transgressions and lies.
For decades, there has been a trend of degrading traditional values and thus
depreciating the words that confer meaning upon them.
In certain Western countries, talking about one’s “nation” or “homeland” raises
suspicions. The ideology that “national” and “social” are opposites tends to be
embraced by public opinion.
Of course, one might well claim the concepts of “homeland” and “nation” have
been exploited and used for self-serving ends by certain people. However, this is not
a reason to consider these concepts dangerous or useless. In a healthy society, you do
not simply throw away the concepts necessary to understand our historical develop-
ment just because someone “polluted” them. You should simply make sure that you
“clean them up,” that you pass them on immaculate in the collective conscience. This
is the modus operandi of societies that protect their legacies that do not allow them to
go to the dogs.
8.6 Depreciation of Values and Decline 133

In the twentieth century, we saw the depreciation of words such as “science,”

“intelligence,” “politics,” and “ideology.” When I was a student in the 1960s, young
people dreamed of becoming scientists because science was placed high on the scale
of values. Science was salvation, it conveyed knowledge and all the innovations that
changed people’s lives. Young people also believed in the value of good; they
dreamed of social justice and of a better future for our country.
This picture faltered and changed after the 1970s. The indiscriminate use of
scientific results, pollution and environmental change, unemployment and poverty
have a direct impact on the decline in the prestige of science.
The degradation of concepts is accompanied by the emergence of new ones.
George Orwell wrote “if thought corrupts language, language can also corrupt
thought.” The syndrome of terminological design is becoming increasingly acute,
where words with a “negative connotation” are replaced by others that are more
neutral or politically correct. A blind person becomes, for the sake of political
correction, “visually impaired” or “sight challenged.”
The “wooden language” syndrome is well known, particularly in politics, which
makes it difficult to find common ground to reach an understanding. This language is
dangerous due to being misleading: saying one thing that might also mean some-
thing else, spreading hollow words, cultivating an ambiguity that is both complacent
and appealing. Alan Greenspan, the former chair of the US Federal Reserve, was not
known for being particularly clear in his speaking, and would ironically say, “I guess
I should warn you, if I turn out to be particularly clear, you've probably misunder-
stood what I’ve said.”
When a concept is lost or replaced by another, the very way in which we can
conceive reality changes. Controlling and shaping language by creating new words
or removing others ultimately makes it possible to control thought. Thus, a “defeat,”
a word with a troubling negative connotation, becomes a “tactical retreat,” or a
“mobile defense war.” During a crisis, we enter a period of “negative growth,”
“economic rigor,” or “economic austerity” leading to “rigor for modernization.”
In recent years, we have seen an attempt at Orwellian redesign in the name of
political correctness. We are told that in order to create a better world free of
inequality and social injustice, we have to “clean up” our historical, scientific, and
technical knowledge from references to individuals whose example and ideas can
hurt and insult people who are discriminated against and live on the margins of
society.
They thus suggest that any reference to giants of science such as Archimedes,
Newton, Schrödinger, or Curie be banned when we teach their results. We are also
told that curricula and the ways they are taught must change with a view toward
more Diversity, Equity, and Inclusion—see [4] for an example.
The reasons for rejection vary either because it is not proper to indicate white
superiority, or because people who lived in a particular place and time, for example,
slave society, colonialist or fascist regimes, revolutions, and world wars, did not
always make the “right choices.”
Whatever the motivations of this movement, naive idealism or political self-
interest, it is a dangerous trade in good sentiments. By purportedly seeking more
134 8 Value Systems and Society

justice, it aims to root out traditional values without actually helping to redress social
injustice. It is downward leveling, a brainwash that wants to change the core values
of our social system, such as individual values and excellence.
Only totalitarian regimes impose censorship and moralism in order to supposedly
protect the common good. Language is man’s most precious common tool for
thinking and communicating.
The redesign or ostracism of concepts, self-censorship, and technology-assisted
language control can imperceptibly change our “thinking software,” the way system
software is changed through over-the-air updates.

8.6.2 The Cycle of Lying

Lying marks the decline of societies and the moral impoverishment of individuals. I
am, of course, talking about damaging lies, not white lies told to avoid offending or
provoking someone.
The most common type of lie is the lie “in word,” i.e., when you consciously say
something different from what you know or believe to be true. You might know that
something is “white” and say that it is “gray” or “black.”
Another type of lie is the lie “in deed,” i.e., when actions fall short of promises.
You may have promised measures that will reduce prices and do exactly the
opposite; you may hold a referendum supposedly to strengthen your country’s
negotiating power, while you are simply holding it to defuse a situation.
However, there is another type of lie that is far more dangerous and devious: the
systemic lie. While the act of lying is conscious in the first two types, in the third type
lying is incorporated, through automatic mechanisms, into the weak, faltering logic
of the person telling the falsehood. This is what we call “fooling oneself.” That is, we
more or less automatically accept as true something which when you examine it
soberly and rationally is completely wrong. It is a lie that charms us and tempts us
toward an unrealistic, utopian impulse. We live through such experiences as children
and adolescents. Considering our wishes to be reality, we create utopias, “myths”
that are convenient to us.
What I want to discuss here is the establishment of this type of lie in our collective
conscience and its disastrous consequences. In a healthy democracy, people telling
falsehoods would be subject to massive and immediate rejection. Decline is inherent
in a remarkable tolerance for lies. There are countless examples demonstrating that
the triumph of the systemic falsehood makes it almost impossible to debate and
criticize based on rational and ethical arguments.
I have explained that many mental functions are automatic. Mental rules and the
ways we apply them are shaped through learning processes that function subcon-
sciously. The way we use concepts determines their meaning. If in practice you court
ambiguity and falsehoods, this becomes second nature. The meaning of the concepts
slacks and the coordinates of thinking become blurred. As a result, the concepts of
words slide away or expand as one sees fit.
8.6 Depreciation of Values and Decline 135

Systemic lies lead directly to relativism, a creeping ideology that reigns supreme
in declining societies. By mistreating and depreciating truth, systemic lies favor
ambiguity, undermine relationships of trust and hinder any development of creative
dialog in societies. I cannot fathom how ambiguity can be creative. It certainly helps
“muddy the waters” and exact unconscionable contracts based on deliberately
ambiguous wording of controversial points.
Some people are fond of the so-called “constructive ambiguity” as a means of
reaching agreement between disagreeing parties by obfuscating thorny aspects of a
debate. However, ambiguity should not be confused with abstraction as discussed in
Sect. 4.1.2. Unlike ambiguity, abstraction is an absolutely creative act, a holistic way
to break down complexity by using a filter. Confusing ambiguity and abstraction is
like considering chicken scratches to be abstract painting.
When a person systemically lies, they consciously create myths and narratives
that explain and justify their attitude. There are numerous examples of this syndrome
of prevarication, which is intended to alleviate guilt and to drape self-interest with a
mantle of superiority. Racism was based on theories of the superiority of certain
races, while the barbarism of the Holy Inquisition invoked the justification that fire
purifies the soul. You justify your intolerant attitude by espousing the theory that you
are an ideologist and pure, and therefore your opponents are corrupt, traitors,
collaborationists, etc.
Systemic lies distort discourse, cutting it off from reality, making it lose its
creative power and role in shaping common visions. It cannot be embodied through
purposeful action. It becomes empty rhetoric, pointless, boring, and even dangerous
at times.
Systemic lies dull the moral sensitivity of people and blind them so that they
cannot see reality soberly. People now resort to “myths” which are increasingly
divorcing them from reality.
Lies become the “bad money that drives out good money,” which ultimately
dominates and excludes any realistic approach.
Let me conclude with a story where a grandchild regularly flatters his old
grandfather, who is moved and gives him some pocket money. One day, the
young boy overdid the praise, so much that his grandfather grumbled and, annoyed,
told him: “I know you’re lying, blast you, but what can I do, I like hearing the damn
lies.” We have learned to live with the lie and cannot do without it.

8.6.3 Self-Convenience and Lack of Self-Esteem

Have you wondered what your life would be like if you had a servant to satisfy your
every whim and desire? If there were no obstacles, no difficulties, if you succeeded
at everything effortlessly, if there was no stopping your appetites?
If humans had what satisfied their appetites automatically and effortlessly, they
would be reduced to “lazy gluttons,” merely a digestive tract.
136 8 Value Systems and Society

If humans were born in a world without adversity, they would certainly be no

more intelligent than plants. Moreover, their happiness would also be lesser to that of
the plants because plants at least struggle to survive.
We live in a period of idolatry. We do not believe in our own forces; we consider
any spiritual search to be unnecessary, we place our hopes for salvation solely in the
material and ephemeral. We see the symptoms of the deterioration of ourselves, of
our self-esteem. We seek outside support, as we do not have moral strength and
fortitude.
The indigenous people once adored the conquistadors because they carried loud
firearms that reminded them of lightning and thunder, which were the hallmarks of
their god. At other times, societies were entranced by cars moving much faster than
humans. Today, we have computers that solve problems more quickly than people.
We are fascinated, without considering that cars, weapons, and computers would not
exist without people, without human creativity.
After the 1970s, societies showed hypersensitivity to what can be considered a
major effort to achieve difficult goals. Everyone talks about protecting everyone’s
rights, about “non-oppression,” about avoiding stress. Few remember that there are
also obligations and that as the saying goes, “no pain, no gain.” As a result, we have
seen a tendency in many European countries to ease educational programs, so that
the “poor children” do not suffer, do not undergo traumatic experiences and, above
all else, know and claim their rights. All this demagoguery is taken place with the
assent of the state, the educational unions, and parents.
Making the bare minimum effort in every task has become the rule, sloppily
patching everything up. We do not care about the future. The quality criteria that
require rigor and precision in carrying out tasks are being slackened.
People only want to hear good news. They close their eyes to the important
problems; they already face plenty, being overwhelmed by their daily complexities,
and the demands of modern life. Let us be positive about everything, without
complaining and caring about civic affairs, which will remain unresolved forever
anyway.
In this downward spiral, only individuals and their wishes in the here and now
count. People think nothing of future generations, they have no vision, they avoid
difficulties and shut their eyes to the problems. This leads to societies that are tired
and expect nothing from tomorrow. In contrast, dynamic societies establish visions
and prepare the future, even if it means facing deprivation.
In the wake of the coronavirus crisis, I have seen how much people’s egos have
grown, especially among young people, who openly and shamelessly put their own
entertainment before public health and the common good.
I have also been amazed by the demagoguery, the pathetic “understanding”
shown by decision-makers and the media: “we understand, they are young people
who want to have fun.” As if it would be such unbearable torment to live through a
summer with no rave parties and all manner of carousing—the way young people of
my generation would spend the summer in the 1960s.
In conclusion, I would say that this leveling of the hierarchy of values is
accompanied by a redefinition of concepts and the relationships between them.
8.6 Depreciation of Values and Decline 137

Even hardened consciences need justification for non-compliance with the rules and
their anti-social behavior. A very frequent attitude is the lack of a sense of respon-
sibility and the excuse that “others are to blame” or it is because of general
circumstances beyond our control.
This attitude is reinforced by the inadequate functioning of institutions, control,
and evaluation mechanisms. It leads to a practical challenge to the “utility” of ethical
principles and discourages efforts to achieve higher goals.

8.6.4 The Triumph of the Idiots

One aspect of social decline is the fact that mediocre personalities prevail in public
life. The question is how social conditions influence and shape this situation.
It is obvious that the dearth of values and favoritism benefit the emergence of the
idiots and discourage the integration of the capable. In a depraved system, the
prevalence of idiots is a certainty. A capable-good individual will avoid such a
system, seeking one that recognizes values, being confident in their abilities. On the
other hand, an individual with reduced faculties of understanding and management
will seek to integrate into a slackened system, which leaves room for irregularities
and is not sufficiently vigilant.
What is an idiot? Someone who is unable to understand reality, to correctly
interpret situations and to link them to actions.
It should be noted that this inability has nothing to do with one’s knowledge and
educational level, or even their IQ.
There are many types of idiots at every level of social organization across all
communities. Useless idiots, the common numbskulls, account for the overwhelm-
ing majority and play a rather useful role in social organization. They serve as a type
of “ballast” that causes friction in the system and makes it less sensitive to sudden
changes.
Here I would like to talk about dangerous idiots who have managed to become
part of a self-sustaining and relatively closed system of elites, who hold a country’s
political, military, and economic leadership positions.
One well-known story is that of Marie Antoinette who, when told that the French
people were starving and had no bread to eat, responded “let them eat cake”!
I also recall a joke about an old, much-hated dictator who was on his deathbed.
Crowds had gathered under his window waiting for his death to be announced in
order to celebrate. Listening to the bustle of the crowd, the dictator was surprised and
asked: “Why are all these people here?”, and received the answer “To bid you
farewell, my General.” And he asked, “Well, where are they off to?”.
Depending on the country and the era, the way in which elites arise varies. In the
past, privilege was almost exclusively hereditary. In contemporary democracies, the
elites are supplied by educational institutions that are open, as a priority, to young
people of top classes—such as Harvard, Columbia, Yale, and Princeton in the USA.
138 8 Value Systems and Society

Of all the types of idiots, the most harmful ones may be those who reached high
places without even realizing what happened to them. When you obtain something
painlessly, without fighting for it, then you are unworthy of respecting it and you
often lose it quite quickly.
Throughout my life, I have had an innate interest in political affairs, particularly
politics in France, where I have lived most of my life and in Greece, due to my
origins. I am struck by the inadequacy of decision-makers at all levels—and
particularly the fact that, despite their proven incompetence in managing public
affairs, they are able not only to survive in public life but also to actually enjoy
brilliant careers, with very few exceptions.
If you are a mismanaging grocer, your business will suffer losses and will soon
drop off the face of the earth. If you are a mismanaging Minister, not only do you
face no risk but, depending on the circumstances, you may actually be promoted.
That is why political parties and governments contain an overwhelming number of
idiots.
Of course, this phenomenon is universal and timeless. It is exacerbated by crises
and occurs in every organization. Even in big technology companies, a good
engineer will not be placed in a managerial position, while those who fail in general
will be promoted; the adage “managers rise to the level of their incompetence” rings
true.
Idiots are mistrustful by nature, especially of those who surpass them intellectu-
ally. When they do not understand something, they entrench themselves in
predetermined views and do not even listen to their interlocutors. This has often
happened to me when I have tried to explain issues that require a degree of
concentration to ignorant managers.
A consequence of mistrust is instinctive cunning. Smart people are not cunning.
They speak and act straight because they have confidence in their judgment and in
their powers.
In order to maintain the acquis, dangerous idiots have certain defense mecha-
nisms. First, they become systematically mistrustful, particularly of what exceeds
their intellect, and especially of intelligent people, whom they avoid like the plague.
An encounter with an intelligent person, however brief, can be very painful to them.
One natural reaction is to reject them one way or another, to entrench themselves in
derogatory or arrogant behavior.
One way in which idiots protect themselves is only enjoying the company of
equivalent idiots, forming cliques of all kinds at work or in social life. A high-
ranking idiot will never hire intelligent advisors. This results in a system that protects
and promotes idiots, while rejecting any element that does not align itself with the
system’s logic.
A typical attitude of ignorant Ministers in the face of a difficult problem involving
political cost is to set up a committee of experts to address the problem. As a rule,
when the Minister’s term of office ends, the committee has not even drawn any
conclusions. There is a famous quote attributed to Talleyrand: “Ne rien faire mais
bien le faire.” It translates to “do nothing, but do it well.”
8.7 About Democracy 139

Unfortunately, the prevalence of idiots is a symptom of the political, economic,

and moral crisis that has plagued post-industrial societies. It undermines one of the
foundations of democracy, which is meritocracy, a topic I will discuss in what
follows.

8.7 About Democracy

The aim of a political system is to ensure economic prosperity and quality of life for
its citizens. The demands for economic growth and social justice must not be made
to compete with each other. Each should assist and complement the other. Social
justice has an ethical, as well as a practical and technical dimension. The absence of
formal and substantive justice is a long-term obstacle to the development of a healthy
economy. When a country has a high rate of unemployment, when a segment of its
population is barely surviving, then it is self-evident that the performance of the
economy cannot be optimal. Democracy without economic progress is a utopia.
Without financial resources, the popular verdict is but an empty shell, political will is
but hot air.
It is common belief that democracy is the system that enables this dual purpose to
be achieved in the best possible way: ensuring equality before the law and enabling
every citizen to achieve self-actualization and, in particular, to create and develop
their personality in conditions of safety and prosperity.
There is no perfect democracy. Despite the common formal characteristics, there
are major differences in its essential modus operandi depending on each society and
country. These differences stem from the degree to which citizens are committed to
common values and, in general, from the effectiveness of institutions.
Democracy is not only formal rules—and this has been categorically proven by
historical experience. It is well known that the popular verdict according to formally
free, successive choices can lead a country to disastrous situations such as the
emergence of Hitler’s and Mussolini’s dictatorships or the deep economic and social
crises of the interwar period.
This system, which was originally conceived and tested in ancient Greece, can be
considered to be optimal, but under certain conditions. These were the social
conditions of the ancient city-states, which formed relatively small and well-
structured groups. Citizens/individuals voluntarily surrendered their personal liberty,
subjecting themselves to the idea of the common good in return for protection and
solidarity. I will not further analyze the relations and links between individuals and
society at large in ancient Greek cities.
Athenian society safeguarded its cohesiveness by directly or indirectly exercising
strict control over citizens whenever they have broken away from certain standards. I
am not referring to illegal acts, but simply behavior, which could, in a way, harm the
cohesiveness of the whole and the principles of the political system—the institution
of ostracism and the condemnation of Socrates are typical of this. It is also well
known that the Athenian democracy did not forgive the mistakes of its leaders.
140 8 Value Systems and Society

The ancient Greek model and the subsequent Roman res publica inspired modern
democracies. Democracy is based on two principles.
One is equality, which takes three forms: isopolity (equal participation in joint
decision-making), isonomy (equality before the law), and isegory (equality in
expressing one’s opinion).
The second equally important principle is meritocracy. That is, the best suited, the
most worthy persons are those appointed to positions of responsibility and decision-
making. Here I must clarify the difference between worthiness and excellence.
Excellence is a distinction in a professional or scientific field. Worthy persons are
those among the excellent who can provide important social or political work and
have demonstrated practical virtue and skills. I believe the distinction is quite clear. It
is a matter of competence, as well as of will to manage and contribute to public
affairs. It is not achieved by obtaining diplomas and qualifications but through
empirical testing. A distinguished scientist or entrepreneur may not be worthy of
contributing to public affairs for reasons beyond their other capabilities, either
because they prefer to be private citizens or because they are not so inclined.
Worthy people are elevated and acclaimed in practice. Of course, it stands to
reason that the worthy would be chosen from among the excellent.
The first references to meritocracy can be found in Plato’s Republic, where he
describes an ideal hierarchical organization of society based on a classification of
citizens. The subject is inexhaustible, and I am neither inclined nor best suited to
make such analyses. I simply want to make it clear that equality alone is not enough
and to stress how important the role of leaders and creators in such a system is.
I have already explained the role of institutions and values in modern societies.
Democracy is the perfect form of government when every individual can responsibly
and honestly manage their freedom—otherwise, it becomes a nightmare. In such
cases, it should be replaced by a more centralized regime where at least decisions and
actions are handled more effectively.
Laws often dictate what is forbidden and set the minimum obligations of indi-
viduals toward their fellow citizens and the state. However—and this cannot but be
the case, unless you have a system where you “watch the watchmen”—it is impos-
sible to control and impose a proper mode of operation when the sense of respon-
sibility has become dull, when consciences have been corrupted.
The first and last bastion of democracy is the spirit of patriotism and the sense of
responsibility among its citizens.

8.7.1 Meritocracy and the Role of Leaders

The application of principles of meritocracy requires the hierarchical structuring of

power systems through institutions with appropriate planning and control
mechanisms.
I would like to stress the importance of a hierarchical structure, which allows for
centralized but efficient information management. For example, there are computer
8.7 About Democracy 141

networks where, for technical reasons, all nodes are equivalent in decision-making;
these are the so-called distributed systems, without a central computer to coordinate
them. In such systems, the computational cost of coordination, for example, in order
to reach consensus, is quite significant in comparison to centralized systems. For
example, distributed systems are used by blockchain technology, which involves
excessive computing and therefore energy costs to complete transactions.
The tumultuous history of revolutions and social change led to the establishment
of commonly accepted democratic principles and structures. These are more or less
well known and have been adopted by forms of government, which, at least in
theory, adhere to the “democratic model.” The problem is the proper application of
these principles. In order to become a successful federation, it is not enough to copy
the Swiss political system. The existence of proper structures is just a prerequisite. It
is ultimately the behavior of citizens that determines the result.
Development and social well-being cannot be achieved without planning and
organization that takes into account the current situation and aims at achieving long-
term goals; in fact, more than goals, what is needed is a vision of what society wants
to achieve and how to live in the distant future. For those who understand how
organized forms of government work, such a vision cannot take shape through light
“coffeehouse” debate.
I have lived in countries where there is a widespread perception that democracy is
thereby guaranteed in abstract terms, through “broad participation,” and through
“collective processes.” If it is true that the essence of democracy is broad consensus
and agreement on a program of actions for the common good, it is not at all self-
evident that this program must be drawn up through procedures involving the “base”
(through general meetings and broad, leaderless working groups). I have seen such
demagogic approaches meet with abject failure.
This idea can only lead to a hodgepodge, a ridiculous situation without logic and
coherence. It often goes hand-in-hand with the idea that every member of the “base”
holds a “piece of power,” thanks to party, trade union, and other connections—and
each person uses this piece of power to contribute to the shaping of choices, while, of
course, defending their own well-intended interests. Thus, goals become an incon-
gruous mishmash, the sum of the aspirations of the many, not a lofty yet realistic
vision that the most worthy have drawn up in a rigorous manner.
In my life, I have had to participate in a large number of committees and councils,
at every level, to formulate policies and evaluate structures. They have cost me
endless hours of drudgery and tedious debates, with results ranging from negligible
to completely counterproductive.
The sweeping concept that does not recognize the necessary role of leaders and
assumes that everyone is equally worthy of governing, regardless of qualifications
and competence, directly affects the very authority of democratic institutions.
Unfortunately, governments are full of ignorant Ministers chosen according to
party and other criteria, even though it is not difficult to find more suitable officials,
even within the ruling party.
Such a concept is a violation of the principle that, in a democracy, all opinions
should count equally when choosing between proposals. However, the process of
142 8 Value Systems and Society

formulating proposals must guarantee that they are technically consistent and con-
tribute to the development of a realistic vision for the common good. Visions cannot
spring forth from endless discussions and upheavals. They are the creations of
inspired minds who take the lead, make a proposal and point the way forward.
They are thought up by people with recognized abilities to create and guide.
I would also like to emphasize that in Greek the word “demiourgos,” which
means “creator,” or “maker” and is the root of the English word demiurge, is a
compound of “δῆμoς,” demos (populace of a democracy as a political unit) and
“ἔργoν,” érgon (work), and initially meant a man who practices a profession. Later,
the word came to mean someone who carries out a “public-interest” work. It was a
title conferred upon officials in many ancient city-states who were responsible for
organizing public affairs. Finally, in Platonic philosophy, the word took on the
meaning of the Maker or Creator of the world.
Of course, this gives rise to the following question: how do leaders emerge? How
do we put the right people in the right position? I would answer that the problem is
not how to find them, but how to convince them to engage with public affairs. There
are worthy people who are willing and able to contribute to the common good.
However, they must be provided with guarantees that their opinion will be respected
and that their name will not be used for self-serving purposes.
It is commonly believed that the wider the process of selecting decision-makers
for a position, the better the result will be. This is another moronic criterion that
ignores the substance, namely the clear definition of the qualifications required and
the fair evaluation of the candidates.
At times, I have conducted rough opinion polls of who are best placed to govern
in Greece and France, two countries I know quite well. Asking people with common
sense from various walks of life who the most capable in a government are, you see
that there is an astounding consensus regarding the persons named. It does not take a
genius to understand that so-and-so is not quick on the uptake, that another person is
susceptible to suggestion, and that yet another has no desire to take risks in order to
do something meaningful. In other words, you need not be clairvoyant in order to
know. One might well ask if it is so simple, why capable and honorable people are
rare in governments?
The answer to this question is also simple. First, no single party consists of
enough people of demonstrable worth to form a good government. Second, by
their very nature, political parties must serve all trends, inevitably including the
ignorant, the corrupt, etc. in their ranks. The leader of the ruling majority often
knows that they do not have the best possible partners, but they do not have a choice.
One way of finding the worthy is by not choosing the most demonstrably
unworthy in the first place, as they do not meet at least one of the two requirements:
integrity and recognized competence. Such a purge makes the game of choice much
simpler. Fortunately, worthy people still exist—but unfortunately, they stay far away
from public affairs.
When I was young and radicalized in France in the early 1970s, I had been
impressed by Charles de Gaulle, despite myopically rejecting him as being “conser-
vative.” The General was famous because he did not pull any punches with the
8.7 About Democracy 143

French. When he addressed them on television, he would call a spade a spade and
ignore the political cost. This man, who led the resistance against the Germans
during World War II, laid the foundations for an independent France by creating a
vision for its decolonization and its development as an industrial and military force.
As a mature researcher, I had the opportunity to learn first-hand, through con-
versations with famous French physicists, how de Gaulle made France a nuclear
power. He secretly had the first atomic bomb built by ignoring not only the reactions
of France’s allies but also the opposition from the majority of the country’s political
establishment. What I mean to say is that the action of a handful of enlightened,
worthy people can change the course of a country’s history.
I want to stress the role of the human factor in creating, envisioning, and planning
the future.
In Western democracies, visions are drawn up by organized groups tasked with
an institutional role (lobbies, think tanks, research institutions, and academies).
There are constants in the policies of major states, regardless of who is in power at
any given time. This is the case with the USA, as well as China, Germany, and many
other major players on the international political scene, such as Israel, India, Japan,
and Korea. We can judge and criticize the way decisions are made, who they benefit,
etc. However, foresight and planning are better than abandonment and inertia.
In a country such as Germany, conflicts die down when the future of the country
is at stake. Let us recall that Schröder’s Social Democratic government took strict
austerity measures in order to address the impending crisis, with no regard for the
political cost.
In Germany, industrial strength is based on sectors such as the automotive
industry, the chemical industry, and production automation. Industry representatives
and research institutions play a key role in identifying priorities per sector. They
work with every German government to develop innovative programs, for example,
for electric cars or new forms of energy. At the same time, there are institutions such
as non-profit organizations, think tanks and experts, as well as press and media
organizations whose words carry weight.
I remember in the past, in France, what effect an article in the newspaper Le
Monde would have when it criticized the government or when it provided guidelines
for the country’s foreign policy.
How do the strategies of the USA for each sector take shape? The system has
organized lobbies that make proposals and at times take initiatives independently of
the federal government. The technological supremacy of the USA and the strategies
to support it are shaped by institutions through group discussions/conflicts where the
military, business, academics, bankers, etc. are represented and are relatively inde-
pendent of the political context at any given time.
I could, of course, also talk about other countries, such as China, which has a
centralized planning model with well-defined priorities that mobilizes all the
resources needed to realize them, aiming at national independence and self-
sustaining growth above all else.
144 8 Value Systems and Society

I will conclude by saying that there is a distorted view of what a democracy is; it
emphasizes the role of the popular verdict but ignores how essential the role of
leaders and creators is in order for the state to function properly.
To paraphrase a well-known aphorism of Albert Einstein, I would say that
“democracy without meritocracy is blind—democracy without equality is lame.”

8.7.2 Corruption and Bureaucracy

The scourge of modern democracies is bureaucracy, a system where the most serious
political and administrative decisions are made by appointed state officials. This
system of administration, which has its roots in ancient Egypt and ancient China,
aims at the self-serving control of administrative structures. It imposes increased
complexity on procedures, which leads to inefficiencies in the management of the
state. Today, it is considered a symptom of ignorance and corruption on the part of
those in power.
Depending on the circumstances, bureaucracy may interfere with government
functions for self-serving purposes or even become dangerous for individual liber-
ties. It can thus cover its impotence to handle public affairs and can organize
networks of corruption for its own benefit.
The biggest problem of an incompetent administrator is how to remain in office
without stirring the pot, making decisions on thorny issues. An administration that
does not want to decide on a problem will look for pretexts by creating structures,
committees that will write long reports. This will prolong the procedures and create
relationships of dependence for those seeking a solution.
In corrupt regimes, a large and poorly organized administration hinders seamless
operation by requiring appropriate tolls at every level of access to services. It is
paradoxical that bureaucracy is often set up on the pretext of ensuring better and
more efficient handling of affairs. However, if we were to believe this argument, we
must realize that even the best decision is useless when it is made too late or when
the cost of the decision-making process is excessive.
In many cases, the setting up of committees—particularly populous and, as they
refer to themselves, “representative” ones, with trade unions, organizations, etc.—is
seen as a guarantee of seriousness and quality. There is a tragic misunderstanding
here about how creative ideas and solutions are born.
Large committees are definitely a problem that negatively affects creativity and
efficiency. While it is right that a proposal, once formulated, be handed over to the
many for criticism, it is not the many who will make a proper proposal. Creation is a
solitary action. Creators need self-concentration to mobilize their ingenuity and
knowledge in order to analyze a problem and come up with solutions.
Bureaucracy usually issues orders, both to experts and non-experts, to come up
with long analyses and reports, purportedly to help it solve a problem. I have seen
such voluminous works that either often draw obvious conclusions or are so poorly
written that no clear conclusions can be drawn at all.
8.7 About Democracy 145

In my career as a researcher, having occasionally taken up administrative posi-

tions to manage research and innovation, I have had the opportunity to study relevant
analyses carried out by well-paid consultancy firms. I generally found them to be
superficial, limiting themselves to describing situations and reaching painless and
obvious conclusions. Analysts will never risk losing a paying customer by voicing
the big truths that could offend their commissioner.
I have experienced many cases of bureaucracy.
One illustrious example of bureaucratic organization is the European Commis-
sion, where the complexity of the structures competes only with the complexity of its
regulations. The perennial aim is, supposedly, to respect the rules and to manage
European funds properly. What everyone disregards, though, is that control mech-
anisms only make sense if the cost of the audit process is commensurable to the
resources being managed.
I have seen endless procedures to audit pennies. There is a different way to
achieve proper management with lower control costs: when the penalties are severe
in any case of violation. This shockingly simple observation eludes the Eurocrats.
What is needed is fewer auditors and tighter control according to simple rules for
meeting substantive obligations and commitments.
We have not yet fully understood the informational nature of human interactions.
I remember in primary school the teacher who taught us the “rule of three” using the
following example for inversely proportional amounts. “It takes 10 people 10 days to
dig a field. How many days does it take 20 people?”. Proportionality implies that
20 people can work in parallel (independently), without each person hindering the
other. It applies when there is perfect allocation of tasks, with minimal interactions.
The same question, when asked with regard to collaborative work such as a
software program to be written by a team of engineers, would make no sense. More
engineers, poorly coordinated, will not shorten but actually lengthen the time needed
for completion [5].
Cooperation involves a cost, and we need to prove that the gains in productivity
and quality are greater than the cost of coordination. Pyramid structures are certainly
more efficient than completely flat ones with no central coordinating body.
If I increase the number of computers I use to solve a problem, the gain in speed
of resolution does not necessarily increase depending on the number of machines. In
fact, beyond a certain threshold, I may not have any increase at all. This is what it
means when we say that we do not have scalability.
What happens with computers also happens with people to a great extent—and
the problem is compounded by the fact that people can demonstrate selfish behaviors
that can make the game of cooperation even more complex.
Those in power have a natural tendency to use even a small portion of that power
to maintain their position within the system. The fight against bureaucracy must be a
constant concern for every democracy.
146 8 Value Systems and Society

References

1. https://en.wikipedia.org/wiki/Precautionary_principle
2. https://en.wikipedia.org/wiki/Boeing_737_MAX_certification
3. Philip Mirowski, Privatizing American Science, 2011, Harvard University Press, ISBN
9780674046467
4. https://poorvucenter.yale.edu/Antiracist-Pedagogy (accessed on 08/01/22)
5. https://en.wikipedia.org/wiki/The_Mythical_Man-Month
Chapter 9
Epilogue

The book is the crystallization of thoughts I have had over many years, spurred on by
an innate curiosity regarding the big questions for which I did not find satisfactory
answers. In this quest, I have been helped by my foundations in applied logic and my
knowledge as a computer scientist and engineer. I have faithfully adhered to
one rule: to distinguish what I can speak about with confidence and to define the
boundaries of knowledge as clearly as possible.
That is why my first concern was to determine which of the big questions can be
studied in a rigorous manner, and which cannot. I soon discovered that there is
indeed a very clear distinction, which arises from the logical nature of the problems,
and which everyone should understand. This would greatly limit vapid discussions,
with the pretext of “scientificity,” concerning ontological or teleological issues.
I therefore aimed at studying gnoseological problems relating to the development
of knowledge—particularly scientific knowledge—and the application of knowl-
edge—particularly technical knowledge. Characterizing knowledge as information
directly connects it to informatics and computers.
I have explained that information is an entity separate from the matter/energy of
the physical world. The proposed framework relies on a dualist vision of the world:
on the one hand, the physical world, and on the other hand, information and
knowledge as a creation of the human mind. This approach transcends the impasses
of an approach that considers matter/energy as the fundamental entity and informa-
tion as an emergent property of matter. How can we comprehend the properties of
our thinking by studying physical phenomena alone?
This outlook helps us understand the close interaction in the development of
scientific and technical knowledge. Unfortunately, we overemphasize the impor-
tance of science while sidelining technical knowledge. It is a shared view that
curricula focus on accumulating knowledge through rote learning, and do not give
due consideration to applications and creativity.
I rank informatics alongside the key fields of knowledge, together with physics,
biology, and mathematics, those fields that provide basic cognitive tools to all other
fields.

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 147
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9_9
148 9 Epilogue

Languages, both natural and non-natural, provide the “toolbox” to describe what
we observe and to understand the world. I have explained that the methods we use to
break down complexity have limits—and that ignorance of those limits leads to
over-optimism and over-expectations. I have analyzed complexity factors for the
predictability of phenomena and the buildability of artifacts. Thanks to the use of
computers, we can overcome certain obstacles and broaden the horizons of
knowledge.
The relationship between natural and artificial intelligence is a very topical issue,
particularly with the discovery and widespread use of learning techniques to develop
smart systems. What will be the equilibrium point between autonomous systems and
human control over their functions in the future? This will depend on technological
progress, as well as the political choices that societies will make in order to avoid a
kind of alienation from decision-making that would result from unrestricted
automation.
Can we fully trust the autonomous management of critical resources and ser-
vices? One thing is for certain: after Homo sapiens, humans who make tools that
multiply their muscular power and create artifacts, we will see a species of humans
who, thanks to the use of computers, will multiply the possibilities to control and
change the world. However, the question is whether this “boom” in intelligence will
be accompanied by a commensurate development of collective consciousness
regarding its proper management.
Informatics provides a basis for comparing natural and artificial intelligence, as
well as possible synergies between them. I have proposed a “mechanistic” view of
consciousness and of the management of freedom, which highlights the complexity
of the problem of understanding mental functions and their simulation by machines.
It leads to a study of human social behavior, as defined not just by material but
also—and mainly—by information and communication factors.
I have tried to demonstrate the informational character of social organizations by
analyzing the mechanisms that shape values, the processes for creating and
reallocating resources, the management and development of knowledge. The decline
of societies can be understood as an enervation of structures and institutions that
allow for a stable, commonly accepted value system and ensure the effective and
secure circulation of information and knowledge.
We can gain a deeper understanding of organizational principles in light of the
theory of computation. A pyramidal organization based on meritocracy is the only
form of organization that allows for effective planning that enables the most worthy,
the creators, to develop visions and proposals for the future and submit them to the
judgment of the public.
Equality takes on its full significance when all citizens have equal opportunities to
make informed choices and equal opportunities to climb up the pyramid. Meritoc-
racy does not cancel the idea of equality but, on the contrary, it makes it more
meaningful.
While writing this book, I adhered to a technical approach and, at the same time,
tried to explain its limitations. I have no doubt that new knowledge will allow people
to make progress in this game of understanding and changing the world.
9 Epilogue 149

However, caution is necessary here. Just as an airplane is not a bird, a computer is

not a human mind.
I find it hard to imagine how creations can exceed the intelligence of their
creators. However, it is possible that the creators end up dominated by their
creations, either due to being unable to manage their complexity or due to intellec-
tual laziness and a propensity for freedom from the burden of choosing responsi-
bly—and that is a terrifying scenario that I would prefer not to imagine at all.
Index

A Artificial intelligence (AI)

Abduction, 41 strong, 1, 3, 67
Abstraction weak, 1
hierarchy, 32, 36, 37, 39, 40 Asimov, I., 81
Action Athens, 21, 130
mandatory, 108 Automation
optional, 93 automated system, 74, 79
prohibitive, 101 Automaton, 54, 55
Actuator, 75, 76, 78 Autonomy
Adaptation/adaptive, 17, 42, 70, 76, 80, 95, autonomous system, viii, ix, 1, 3, 70, 71,
122, 123, 130 74–80, 83, 84, 91, 95, 123, 148
Administration/administrator, 144 autonomous transport system, 26
AI, see Artificial intelligence (AI) autonomous vehicle, 76
Air traffic control, 26 Autopilot, 76, 80
Alexander the Great, 131 Awareness
Algorithm, 14–17, 24, 32, 35, 53, 54, 61, 66, self-awareness, 95, 116
90, 111
Alibaba, 80
Amazon, 129 B
Ambiguity, 36, 132–135 Bank
Anomalies banking institutions, 129
time, 64 Behaviour, 28
Appetitive part, 96 Big Bang, 25, 41, 89–91
Apple, 13, 80, 129 Biology
Application biological systems, 38, 42
critical, 60, 121 Blockchain, 129, 141
of knowledge, 3, 31–49, 147 Boeing 737-Max, 123
Risk Management Principles and Their Boiling frog, 84
Application, 44, 121–123 Brain
Archimedes, 29, 133 Brain-Computer Interaction, 91
Architecture, 9, 76, 77, 79, 91, 95, 96 human, 40, 81
Argentina, 132 mind-brain relationship, 90–91
Aristotle, 8, 9, 32, 124 Bureaucracy
Arithmetic, 20, 33, 35, 53, 61 corruption and, 144

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2022 151
J. Sifakis, Understanding and Changing the World,
https://doi.org/10.1007/978-981-19-1932-9
152 Index

C Computing
Calculus about, 12, 31, 64
differential, 22, 35, 61 knowledge and intelligence, 85
propositional, 34 natural, 17
Car scientific knowledge and, 53–55
self-driving, viii, 26, 74, 76, 80 system, 26, 36, 57, 63, 64, 81, 91–101
Carter, J., 126 Concept
Century degradation of, 132–134
20th century, 9, 12, 34, 35, 84, 111, 120, Condition
127, 132, 133 environmental, 98
21th century, 70 initial, 23, 106
Certification living, 96
self-certification, 80, 123 material, 132
Challenge mental, 104
scientific and technical, 46 physical, 95
Charity, 94, 100 political, 130
China safety, 104, 105
Chinese Room Argument, 70 social, 130, 137
Chinese virus, 122 Conflict
Christianity/Christian, 8, 126 conflict-free state, 105
Circuit cooperation-conflict model, 128
electrical, 37, 38, 40, 55 resolution, 60, 124
integrated, 11, 36 Conscience
neural, 55, 89 collective, 110, 134
Cleobulus of Lindos, 124 Conscious, 1, 8, 65, 66, 90, 95, 96, 101, 103,
Club of Rome, 43 108–114, 119, 126, 134
Cognition, 12, 32, 89 Consciousness
Command, 60, 62, 66, 75, 76, 78 collective, 2, 110, 118, 148
Completeness, 26 components of consciousness, 94–96
Complexity and computers, 60
cognitive, 68–69, 73, 107 Continuum/continuity, 55–57
computational, 23, 24 Control
conceptual, 26, 27 controllable action, 99, 100, 103
disorganised, 39, 118 non-controllable action, 100, 104
epistemic, 22, 23 Conway, J.H., 55
organized, 39 Coronavirus
of synthesis, 27 crisis, 44, 123, 136
of validation, 27 pandemics, 121, 122
Component, 27, 31, 32, 37–39, 42, 63, 64, 73, Correctness
82, 94–96 and erroneousness, 120–121
Computation political, 133
computational complexity, 16 Corruption
computational processes, 53 and bureaucracy, 144–145
theory of, 29 Creation
Computer learning and, 111
conventional, 17 Creator, 8, 112–114, 140, 142, 144, 148, 149
natural, 17, 55 Criticality
supercomputer, 19, 74 critical mass, 48, 49
system, 11, 27, 40, 59, 62–64, 74, 82, 91, critical system, 80
121, 141 Culture, 48, 56, 112, 130
world, 14, 16, 63 Cuomo, A., 123
Index 153

Curie, M., 133 Evaluation, 137, 142

Cyber-attack, 82, 129 Evil, 81, 83, 125, 126
Evolution, 10, 32, 33, 41, 42, 55, 60, 62, 69, 82,
89, 91, 105, 110, 117, 129
D Excellence, 48, 49, 93, 94, 97, 134, 140
Darwin, C., 41, 42 Existentialism/existentialist, 106
Dawkins, R., 12 Experience
Deadlock, 59, 60, 64, 104–106 transcendental, 109
Decision Experiment
decision-making, 108, 144 thought, 71
Deduction, 34, 41 Experimentation, 20, 29, 62, 92
de Gaulle, C., 142, 143 Experimenter, 22, 72, 73
Demagogy/demagoguery, 126, 136, 141 Explanation
Democracy, 3, 44, 118, 120, 123, 130, 134, explainable artificial intelligence, 85
137, 139–145
Democritus, 8, 32, 37, 56
Depreciation F
of values, 131–139 Facebook, 83, 129
Design Faith
model-based, 62 and dogmatism, 127–128
models, 26, 27, 62, 80 False/falsehood, 9, 10, 34, 35, 118–120, 124,
space, 27 134
Determinism/non-determinism, 8, 49, 61, 64, Fear/fearless, 96, 97, 99, 117, 120, 122, 123,
106, 132 125
Dijkstra, E.W., 11 Feeling
Dogmatism active, 97
faith and, 127 negative, 98, 101
Dynamics, 21, 24, 40, 56, 57, 66–68, 78, 79, passive, 97
103, 110, 115, 118, 128, 130, 136 positive, 98, 101
Dynamic system, 53, 57, 67, 92 Feynman, R., 62
Formalization, 26, 33, 34, 73, 90
Fortran, 82
E Framework
Economics/economy/economist, 8, 14, 15, 23, regulatory, 84, 85, 117, 129
24, 28, 31, 33, 38, 39, 42–46, 48, 49, 57, France, 123, 138, 142, 143
58, 60, 64, 68, 79–81, 83, 92, 94, 95, 99, Freedom, 32, 46, 64, 91–93, 95, 100, 105–108,
102, 103, 118, 119, 122, 123, 128–133, 114, 116, 124, 128, 140, 148, 149
137, 139 Free will, 105–108
Education Function
educational programs, 136 mathematical, 15, 73, 90
Efficiency, 49, 62, 74, 81, 84, 111, 130, 144
Einstein, A., 63, 69, 144
Electro-mechanical system, 22, 64 G
Emotion, 96, 97, 110, 120 Galileo, 29, 72, 119
Environment Gates, B., 81
external, 67, 76–78, 94, 95, 99 Germany, 43, 49, 143
internal, 67, 74, 76–78, 94, 95, 99 Gnoseology
physical, 76, 95, 96, 110 the gnoseological “how?”, 10
Epilogue, 3, 147–149 gnoseological value, 93, 118–124
Epistemic values, 118–120 Goal
Equation management, 78, 79, 102–108
system of, 22, 38 negative, 102
Erroneousness positive, 102
correctness and, 120–121 Gödel, K., 15, 16, 20, 35, 57, 119
Ethics Good and evil, 123–127
ethical values, 124 Google, 69, 74, 80, 83, 129
154 Index

Government, 83, 123, 138, 140–144 Intel, Mobileye, 80

Grammar, 42 Interaction
Greece, 32, 43, 45, 138, 139, 142 Brain Computer Interaction, 91
Greek, 2, 8, 13, 29, 33, 45, 113, 114, 139, 140, human, 96
142 Internet
Greenspan, A., 133 Internet of Things, 1, 74, 79, 84
protocols, 82
Introduction, 1–3, 29, 36
H Invariant, 54, 63
Happiness Isegory, 140
collective, 115, 125 Isonomy, 140
individual, 115, 125 Isopolity, 140
seeking, 115–116 Israel, 49, 143
Hawking, S., 12, 25, 81
Hayek, F.A. von, 39, 130
Heraclitus, 9 J
Hero/heroism, 97, 101 Jew, 126
Heron of Alexandria, 29 Judaism, 8
Hierarchy Justice, 93, 100, 126, 133, 139
abstractions, 32, 40
of values, 136
Hitler, 139 K
Homo sapiens, 148 Kant, E., 115
Huawei, 80 Knowledge
Human boundaries of, 23, 28, 46, 147
factor, 42, 43, 49, 65, 74, 118, 123, 143 empirical, 1, 7, 19–21, 28, 34, 53, 85
intelligence, 1, 3, 65–85 mathematical, 1, 7, 14, 18, 20, 21, 25, 28,
society, 2, 105, 117–118, 124, 131, 136 29, 31, 34, 35, 42, 73, 80
meta-knowledge, 1, 21, 29, 104
a priori, 29, 53, 75, 106, 127
I scientific, 1, 2, 7, 19–25, 27–29, 34, 46,
Idealism, 133 53–55, 71, 72, 80, 85, 91, 92, 127, 133,
Ideology, 123, 127, 132, 133, 135 147
Idiot self-knowledge, viii, 90
the triumph of idiots, 137–139 technical, 1, 2, 7, 19, 20, 25–29, 42, 46, 102,
Informatics, 2, 3, 11–29, 31, 36, 38, 43, 53, 69, 133, 147, 148
89, 90, 147, 148 Korea, 143
Information Kurzweil, R., 81
syntactic, 14
Infrastructure, 48, 76, 82–84
Institution L
banking, 129 Language
research, 46, 48, 49, 143 and consciousness, 19
the role of, 3, 85, 128, 130–131, 140 decay of, 132–134
Intelligence formal, 25, 34–36, 57, 67, 73, 90
computing, knowledge and intelligence, 53 natural, 13, 32, 89, 90
human vs. artificial Intelligence, 1, 65–85 programming, 13, 63, 70
human intelligence, 65–70, 81, 117 wooden, 133
(see also Artificial intelligence (AI)) Latin American, 130
Index 155

Law mental, 22, 76, 95, 96, 98, 102, 115

and Commandments, 126 model-based design, 62
ethical, 107 Modelling
Mosaic, 124 tools, 23
physical, 35, 55 Modularity
of physics, 10, 14, 21, 25, 63, 69, 92 assumption, 37, 38
Leader, 122, 139–142, 144 Money
Learning monetary values, 93, 100, 101
artificial, 17, 65, 70, 148 Morality
machine, 1, 17, 55, 65, 70–73, 80 moral crisis, 120, 139
Leibniz, G.W. von, 61 moral value, 115
Leonardo da Vinci, 29 Mosaic Law, 124
Liberty, 44, 84, 101, 102, 104–105, 119–121, Musk, E., 81
127, 131, 139, 144 Mussolini, 139
Lie Myth, 32, 92, 134, 135
systemic, 134
Logic, 1, 2, 9, 25, 31, 32, 34, 35, 57, 59, 66,
111, 116, 120, 134, 138, 141, 147 N
Lying Nation, 129–132
the Cycle of Lying, 134–135 Neo-liberalism, 129, 130
Network
neural, 66, 72, 73
M Newton, Newtonian mechanics, 39
Machine Nordic countries, 49
computing, 54
learning, 1, 17, 55, 65, 70–73, 80
Management O
goal, 77, 78, 96, 102–108 Observation/observer, 10, 20, 25, 28, 29, 35,
and planning, 77, 102–108 36, 41, 54, 56, 61, 63, 68, 72, 91, 145
risk, 44, 83, 121–123 Ontological
Marie Antoinette, 137 the ontological “what”, 9–10
Market, 48, 84, 100, 122, 123, 128–130 Optimization, 78, 103, 122
Materialism, 8, 130 Organization
Mathematics social, 82, 92, 94, 105, 108, 117, 118, 123,
and logic, 20, 25, 29, 34, 90 128, 132, 137, 148
mathematical induction, 57 political, 132
Mayan calendar, 81 Old Testament, 124, 125
Memory, 12, 14–16, 36, 39, 54, 57–60, 63, 64, Orwell, G., Orwellian, 133
69, 102
Mental system/mental phenomena, 10, 19, 89,
90, 95–102, 104, 106, 109, 117 P
Meritocracy Parmenides, 9
and the role of leaders, 140–144 Patriotism, 94, 140
Method/methodology/methodological, 2, 7, 8, Perception, 10, 12, 67, 77–80, 95, 109, 141
12, 23, 24, 26, 28–29, 36, 38, 41, 44, Performance, 24, 27, 64, 93, 108, 110, 112, 139
53–54, 71, 72, 80, 85, 111, 112 Phenomena
Microsoft, 129 mental, 10, 19–21, 32, 40, 57, 60, 89, 90,
Middle Ages, 119, 124, 128 106
Mind natural, 32, 55, 61, 148
the Mind as a Computational System model, physical, 10, 20, 23, 24, 29, 32, 53–64, 66,
91–101 90, 106, 147
mind-brain relationship, 90–91 social, 10, 23, 24, 42, 57, 60, 68, 105, 119,
Model 130
computational, 1, 17, 23, 62, 66 Philosophy
mathematical, 11, 22, 73 of the individual, 8
156 Index

Physics Researcher, 43, 45–49, 111, 112, 114, 143, 145

classical, 56 Resource
digital, 54 consumable, 59
laws of, 21, 63, 69, 92 non-consumable, 58
quantum, 2, 38, 62 qualitative, 58
Planck length, 36, 56 Responsibility, 46, 83, 85, 97, 106, 119, 121,
Planning, 48, 78, 79, 96, 102, 103, 131, 132, 124, 125, 131, 137, 140
140, 141, 143, 148 Res publica, 131, 140
Plato Revolution, 11, 47, 83, 129, 132, 133, 141
Plato’s Republic, 140 Risk
Point of equilibrium, 10, 98, 101 management, 44, 82–83, 85, 121–123
Policy, 43, 45, 49, 80, 123, 129, 141, 143 Robot
Post-truth, 120 chess-playing, 75
Precautionary principle, 122, 123 football-playing, 75
Precondition, 57–59, 99, 100, 107 Roman Empire, 131
Prediction Rule
without understanding, 73–74 ethical, viii, 32, 94, 108, 121, 124, 125
Privacy, 84 golden, 111, 124
Problem-solving, 21, 58, 111 induction, 35
Process inference, 35
computational, 55 mandatory, 108, 124
physical, 29, 54, 55, 63, 72 proscriptive, 108, 124
Processor, 58, 89, 109, 130 rule of law, 128
Program rule of three, 145
programming language, 15, 54 traffic, 93, 94
Property, 2, 14, 19, 26, 27, 32–34, 38–40, 42, Russel, B., 25, 35
45, 55–57, 61, 63, 64, 66, 68, 77, 91, 99, Russia, 129
118, 147
Proposition
undecidable, 35, 119 S
Pythagorean, 32 Safety, 27, 46, 75, 76, 78, 80, 82–83, 85, 93, 94,
96, 97, 104–106, 121, 123, 139
Schröder, G., 143
R Schrödinger, E., 133
Racism, 135 Science
Ramanujan, S., 113 computer, 11, 54
Rational part, 96 fiction, 81
Reaction, 58, 75, 76, 96, 102, 103, 109, 110, natural, 40
138, 143 A New Kind of Science, 54
Reagan, R., 47 physical, 28, 54
Real-time, 55, 57, 65, 66, 76, 78, 79, 81 social, 24, 38
Reasoning systems, 43
common sense, 68 Scientificity
rules of, 66 the limits of scientificity, 40–42
Reflection, 77–79, 95 Searle, J.R., 10, 70
Regime, 48, 133, 134, 140, 144 Security, 82–83, 93, 96, 121, 123
Relationship Self
cause-and-effect, 41 self-awareness, 116
Renaissance, 29, 33 self-censorship, 134
Repository, 77, 78 self-certification, 80, 123
Requirement, 26, 27, 36, 78, 99, 101, 142 self-confidence, 97
Research, 17, 35, 40, 41, 43, 45–49, 67, 69, 84, self-control, 110
89–91, 108, 126, 143, 145 self-convenience, 135–137
Index 157

self-discipline, 125 computing, 62, 92

self-driving car, viii, 26, 74, 76, 80 discrete, 59
self-esteem, 97, 100, 125, 135–137 dynamics, 22, 62, 63, 77
self-fulfillment, 96, 97 electro-mechanical, 60, 63
self-interest, 135 neural, 17, 66, 80
self-knowledge, viii, 90 value, 3, 58, 62, 92, 94, 98, 99, 101
self-learning, 77–79, 98, 103
self-preservation, 117
self-sacrifice, 93 T
self-serving, 132, 137, 144 Tactic, 75, 76, 78, 79, 102–105, 121
self-sustaining, 137 Talleyrand, 138
Selfish/selfless, 101, 117, 125, 145 Technical values, 118, 120–123, 128
Semantics Technology
model, 67, 90 technological singularity, 81
network, 67 Telecommunications, 11, 81, 83, 129
relationship, vii, 13, 89, 90 Teleology
Sensor the teleological “why?”, 8–9
sensory information, 67, 76, 77, 79 Tesla, 80
Shannon, C., 15 Thales, 22
Simulation Theorem
physics, 62 Fermat’s last, 119
program, 62 Gödel’s, 20, 35, 57
technique, 27, 29 Pythagorean, 20
Society Theory
human, 2, 105, 117–118, 124, 131 atomic, 32
slave, 133 of computation, 15, 29, 53, 61, 66, 69, 111,
value systems and, 117–145 148
Socrates, 34, 113, 114, 139 relativity, 25, 29, 36, 119
Specification, 26, 27, 121 scientific, 1, 29, 35, 53
Spirit Thermostat, 58, 74–76, 79
spiritual value, 127, 128 Theseus, 126
Spirited part, 96 Thinking
Star Wars, 47 analytical, 113
State automatic, 65, 66, 109, 112
kinetic, 77, 78 conscious, 65, 66, 90, 95, 109, 110, 112
variable, 35, 57, 58, 62, 78, 98, 99 fast, 65, 66
Strategic Defense Initiative, 47 slow, 65
Strategy, 67, 76, 78, 79, 102–104, 107, 115, synthetic, 25
143 Time
Stratification, 32, 36, 37, 39, 128 endogenous, 62
Subconscious execution, 14, 58–60, 62, 64, 99
selves, 109–114 exogenous, 61
Swiss political system, 141 logical, 53, 62
Symbol physical, 14, 33, 55, 58, 59, 61, 62, 93
symbolic language, 13 real-time, 78
Synchronization space-time, 14, 53, 55
in nature, 61 Transaction, 33, 58, 59, 102, 125, 129, 141
Syntax Transcendental experiences, 110
syntactic information, 14, 15 True/truth/truthful/truthfulness, 1, 2, 9, 10, 12,
System 18–20, 23, 25, 27–29, 34, 35, 38, 41, 54,
biological, 39 63, 64, 66, 99, 104, 105, 111, 113, 116,
complex, 22, 26, 27, 38, 62, 63, 75, 76, 78, 118–121, 130, 134, 135, 138
84 Trump, D., 122, 123
158 Index

Trust/mistrust, 42, 71, 79–80, 84, 125, 129, military, 93

132, 135, 138, 148 monetary, 93, 100
Turing, A., Turing test, 70–71 moral, 93, 94, 115
Turkey political, 93
“inductivist turkey”, 25 religious, 93
Twitter, 83 scale of, 118, 133
system, 2, 92–94, 99, 101, 106, 107, 131,
148
U Vehicle
Uber, 80 autonomous, 76
Understanding Verification
the boundaries of, 89 mathematical, 54
prediction without, 73 Virtue, 94, 125, 127, 140
Unemployment, 82, 133, 139 Vision, 1, 3, 19, 46, 47, 49, 79, 84, 111, 114,
United Nations, 131 120, 130–132, 135, 136, 141–143, 147,
University, 46, 48, 107, 111 148
USA/US/US Federal Reserve, 47–49, 80, 122,
123, 126, 130, 133, 137, 143
Utopia, 134, 139 W
Waymo, 17, 80
Wiles, A., 119
V Wittgenstein, L., 32
Validation Wolfram, S., 54
complexity, 24 World
experimental, 7, 27, 40 computer, 7, 11, 14, 16, 63, 111
technique, 27 cyber-world, 36, 83
Validity physical, 10, 21, 28, 29, 33, 36–38, 53, 54,
degree of, 1, 118 56, 63–64, 90, 95
Value Western, 129
common scale of, 99 World Health Organization, 131
common value system, 2, 92, 94, 118, 124, World War II, 48, 143
130, 131
discrete, 53
educational, 93 Z
individual scale of, 95, 96, 98, 127 Zeno, Zeno’s paradox, 32
legal, 93