UNIT

INTRODUCTION
TO AI
Artificial Intelligence is
understood differently
by different people, yet the term along with the related
terms like machine learning, big data, neural
networks has become a standard item of
public and private arena. Suddenly, it seems that discussion in the
development of artificial everybody everywhere is talking about this field. The
intelligence
or
simply Al isrelatively new.
Five to six decades ago, the existence of artificial
But the
intelligence was limited the
realm of science fiction.
to
development of science and
technology during the interim period changed everything. Various
scientists started thinking seriously about the
possibility of machines during the 1940s. The term artificial
intelligence was first used in 1956 at the Dartmouth
Conference by John McCarthy. During the early
period of development, the Al was restricted to
problem-solving.
The focus was on using symbolic methods for developing Al systems. The US
of the Department of Defence was
one
pioneers of Al research during this period. Even back then, this
in
creating department was interested
computer that acted and behaved like human beings. The human like robots and
a

intelligence are still restricted to science fiction, although research related to them have been artificial
for a considerable time now. We will look at going on
the current approach towards
human brain and the developing systems mimicking
problems associated with it in the third chapter.
Al CHANGING THE HUMAN WORLD.

Emotion Behaviour Attribute Object

Vernie.Thorpe @Venie Thorpe

I Love my new IPhone. Im u sing the new lunes app to
sream the new laylor Suwift album. Thanks Applel

Hashtag People Drand

ARTIFICIAL INTELLIGENCE I Kips (125

 across the world are using it to
of making human tasks easier. Organisations
Al technology is capable tasks. These systems and/or machines
are performing specific
create artificial systems and machines that
efficiency.
are reducing time effects, and errors while increasing

BUT NOT WITHOUT RISKS

benefits. This is true for Al as
risks, along with
Every technological advancement and breakthrough brings as more efficient and at the
well. Al can also be misused. It can make weapons, which can act
be used to

use this technology

to maximise the impact of their attacks.
same time destructive too. Terrorists can
also poses serious privacy concerns. Due to
Cybercriminals can use it for hacking people. Al technology
for government regulations in this sector.
this, organisations and scientists across the world are pushing
fiction for years. Everyone talks about
Artificial intelligence has been a substance in legends and science
this technology. Many doomsday prophets have even predicted a scenario in which the Al will take over,

and human beings will become slaves.

ARTIFICIAL INTELLIGENCE HISTORY

The term Artuficial Intelligence or Al was first used in 1956, but it did not gain popularity untl much
later. With dime, the use of computers increased, the internet was born, and the mobile internet picked
pace. All of this led to an increase in data volumes, the improvement in the computer's computing power,
increased data storage capabilities, and as a result the Al gained popularity.

The research in Al during the 1950s was restricted to problem-solving and symbolic methods. During
1960s, the pioneering work in the Al field was done by the US Department of Defence that was working
on creating computers capable of acting like human beings. In 1970s, the US Defense Advanced Research
Projects Agency (DARPA) completed the street mapping project. Incidentally, the first intelligent personal
assistant was produced in 2003, Iong before Siri, Alexa, or Cortana.

Human like robots and artificial inte!ligence of science fiction are far away from reality. The development
of self-aware Als has not shown much progress, and it will continue to exist in the field of science fiction
for some more time to come.

A Brief History of Al
Artificial intelligence was first conceptualised in 19U by Alan luring and has seen a number of innovations
and false starts since then. In the past three or so years, innovotion has accelerated greatly. Following is O
seiection of some of the major milestoneS in Al during the past seven decades.

950-Alan Turing 1959-Artificiol 1974-The 1987- 1994-First 2002-Amazon 2011-Apple

Computing Intelligence first Al winter; Second weD search
Al winter
replaces human releases dr,
machinery and Loborotory funding engines editors with an a personal
intelligence, asks founded at MIT and interest begins launchedd automated voice agent
Can machines evaporate
system
think?
******************

L O
1956-Artificioi 1975-MYCIN, a system 1989-NASAs 1997-1BMs Deep 2016-Coogle's
Inteligence' coined that diagnoses bacterial AutoClass program Blue beats world AlphaCo defeats
John McCorthy ot iníections and recommends Used to discover Lee Sedol; one of
by champion Garry
Darimouth College antibiotics, is developed new classes of stars
Kasparov at chess the world's leading
Go players

126 KipsI ARTIFICIALINTELLIGENCE

Al ETHICS
The term 'AI Ethics' is a blanket term that is used to deal with all the ethical concerns and issues related
to Al systems. Al ethics are normally divided into two categories: Concerns related to data (Bias and
inclusions) used in the Al systems and the implications of the Al technology itself.

In the most straightforward way possible, we can define 'ethics' as a system of moral principles that govern
individual's behaviour or actions. Ethics are concerned with what is suitable for individuals and societies.
Similarly, ethical concerns are the issues, situations, or concerns that cause individuals, societies and/or
organisations to evaluate different choices in terms of what is right (ethical) and wrong (unethical).
On one end of the spectrum, we have concerns related to the use of Al to control defence equipment.
For example, Al-controlled nuclear missiles capable of making independent decisions or Al equipped
armed military drones. COn the other end, we have considerations regarding human cost of adoption for
Al technology. And then, there is a third side of the concerns that is related to the data used in the Al
systems, and the manner in which the Al makes decision. For example, can Al system make unbiased
decisions when human data on which the system is trained always has biases?
The real-life incident mentioned above is only the tip of an iceberg. Earlier, we learned about Al and
its components, how these components relate to our daily lives, how they relate to the sustainable
development of our planet and the possibilities associated with this technology.

ARTIFICIAL INTELLIGENCE
e
I Kips (127)
EXCITE
INTELLIGENCE
MEANING OFARTIFICIAL
o v e r the years
have added different meanings
is
easy to define. Scientists
not
The term artificial intelligence number of related terms
become complicated due to the presence of a
further
to the word. The issue has in fact not the s a m e .
used interchangeably with Al when they are
often
and technologies, which are

DEFINITION hardware and

i.e.,
artificial intelligence can be defined as the ability of computer systems,
In simple words, to use intelligence.
software, to do tasks that normally require human beings
Intelligence. Almost all of the Al
definition of Artificial
This definition describes the commonly accepted which
definition of the term intelligence,
with it. But this definition leaves out the
scientists will agree
we have to
understand the usage and
difference of opinion in the field. For understanding this,
causes a

human intelligence.
meaning of the term

refers to one o r more of

when used in connection to
hunman beings,
The term intelligence,
the following behaviours:
Reasoning Perception
Planning Learning
Manipulation Creativity
Problem solving Motion

Social intelligence Knowledge representation

etc. all fall under the term inteligent systems or
smart navigation systems,
Language recognition systems, from experience, then he is known as intelligent.
artificial intelligence. If a
human is capable of learning
will fall under artificial intelligence.
are capable of learning
from the data fed into them
that
Similarly, systems
INTELLIGENCE
RELATED TO ARTIFICIAL
FIELDS CLOSELY artificial
that are closely related to
fields, technologies, methodologies, techniques, etc., is magnified
There are many and the problem
with Artificial Intelligence,
These fields are often confused Some of these
intelligence. terms interchangeably
with Artificial Intelligence.
of these
when experts use many
terms are:
created system to
refers to the capability of the artificially
Machine Learning: Machine-learning which a natural language
processing data. For example, the process by
learn from experience, i.e., by m o r e . This is one of
the subfields of Artificial
itself when used m o r e and
improves
translation program

Intelligence. methods of modelling data within the discipline

These are one of the techniques or
brain. The
Neural Networks: stimulate the working of the human
This method tries to mimic or
of Artificial Intelligence. machine learning, i.e., they improve
with the usage. We
this method are capable of
networks created by
3.
about these networks in Chapter
will learn m o r e
data where all artificial intelligence
A discipline related to collecting, managing, and analysing science
Data Science:
a number of areas in
artificial intelligence in which data
on data. There are
systems are dependent
include:
contributes and these

INTELLIGENCE
ARTIFICIAL
128 Kips I
 Managing Big Data

Data Acquisitions

DataModelling
DataAnalysis
Data Curation
Robotics: Robots are computing machines that have both hardware and software components and robotics
is the field that deals with construction and programming of robots. Artificial intelligence contributes
towards the software components of robots which can behave and work like a human
being. For example,
computer vision helps robots to identify their surroundings. Natural language processing helps them to
communicate more effectively. Neural networks allow them to mimic emotions. Machine learning helps
chem to learn.

IMPORTANT PHASES IN DEVELOPMENT OF ARTIFICIAL INTELLIGENCE

As we have learned, the shifting of the artificial intelligence from science fiction to real science started in
1940s, and the term artificial intelligence was not developed until 1956. Thereafter, the pace of the
development of the Al field was not spontaneous. Instead, it took its time and happened in bouts. The
phases in the development of artificial intelligence have been divided into four distinct waves:

FIRST WAVE AND SECOND WAVE

The first serious boost to the
development
of the artificial
intelligence happened with
the growth of the
internet. This development Fourth wave
Al and robotics
bOOsted the generation and the use of data
at a rate that was never seen in the
history Number of users or
of the
world devices: over 50 billion
previously. This is known as the
first wave of
artificial intelligence. Third wave (2020)
The Internet of Things
second substantial boost to the Number of users oor
development of artificial intelligence devices: 10 billion
Ppened with the advancement of mobile
Second wave (2010)
nernet. This development shifted the focus Mobile Internet
rom static
(and often outdated) data to the
Number of users or
eal-time data. This brought a fundamenta devices I billion
change in the manner in which the
First wave (2000)
ystems
Ihis is
were developed and
operated. Internet
known as the second wave of Al.
THIRD WAVE
The third serious
herious boost to the development of Artificial Intelligence happened ich the evolution of
"Internet of
of Things.
Things. The Internet of Things calls for all the appliances and sensors in the world to

ARTIFICIAL INTELLIGENCE | KipS 129

become smarter, ie., to be capable of communication. These interconnected and communicating devices
increased the number of the internet-connected users or devices from 10 bllion in the second wave to 50
billion. This fivefold increase led to an unprecedented jump in the avallability of data due to the multiplier
is known as the third wave.
effect. This change in the working and functioning of the artificial systems
FOURTH WAVE
wave. In this wave, true artificial
Many sclentists and philosophers consider this as the next coming
what human bengs are capable of doing
intelligence will emerge. This artificial intelligence will surpass
today.

WORKSHEET 1: Al-related fields

all that apply, eg, ML
In the following quiz, try to determine which Al-related fields are involved. You have
to select

involves some kind of statistics. (First one has been done for you as a sample.)

Driverless cars
STATISTICS

Maneuvering of a rocket in
its orbit in the space

ROBOTICS
Optimisation mechanism
for onlne advertisement
optimisation

MACHINE
Chat-bots on the website LEARNINGG
for customer service

Find out what do they do and which company

Consider the following smart products.
produced them.
What it does Company
Product
Chromecast

EchoDot
Harmony Elite

Hue
Nest Hub

Smart Things

130 Kips I ARTIFICIAL INTELLIGENCE

PES OFARTIFICIAL INTELLIGENCE
hroad level. there are two
types of Artificial Intelligence:
ARTIFICIAL NARROW INTELLIGENCE (ANI)
An existing type of Al that is capable of performing
fof these systems is narrow in the sense that they aresimple tasks requires basic intelligence. The intelligence
eases, they can pertorm this one task better than
only capable of performing only one task, but in many
human beings.
Eor example, Google Now and Alexa can
recognise and speak multiple languages. The number of languages
sooken and understood by these systems are more than what humans
can speak and understand. However,
Google Now and Alexa cannot drive vehicles, but in the future
the Al. which is capable of driving vehicles. they might be able to communicate with

WHAT CAN ARTIFICIAL NARROW INTELLIGENCE DOo?

Artificial Narrow Intelligence is capable of undertakingg a number of actions, which include the
following:
Processing videos for understanding them. For example, recognising when an intruder enters a house
or when small children are in
danger in the house.
Performing routine of everyday tasks, which might require some limited decision
ordering milk, switching off the power in empty rooms, switching on Air Conditionersmaking. For example,
when needed.
Performing simple tasks done by secretaries. For example, taking dictations, scheduling
taking minutes of meetings, reading emails, etc. appointments,
Performing tasks that require processing of data. For example, assisting doctors in diagnosing diseases
based on symptoms.

ARTIFICIAL GENERAL INTELLIGENCE (AGI)

Artificial General Intelligence refers to the science fiction idea of artificial
intelligence, i.e., the computer
Sstems, which are at least as intelligent as human beings, if not more. These systems will be capable of
carning and developing themselves. They will not be limited by the tasks that they can perform. These
systems are in the research stage, and much breakthrough has not been made regarding them.

WEAK VERSUS STRONG ARTIFICIAL INTELLIGENCE

ruicial intelligence systems can also be classified into strong Al and weak Al systems:

Weak Al: It is the intelligence that mimics intelligence but does not process it or in other words,
acts intelligently without having any intelligence. All of the Al systems in existence today fall under this
category. For example, a self-driving car might recognise that it is raining, but will not understand what
rain is. A smart home security system can recognise intruders without understanding what an intruder is.

erong Al: Strong AI systems are theoretical systems that actually process intelligence. These systems
Currently are in the research stage.

DISCIPLINE OF ARTIFICIAL INTELLIGENCE

have seen, the or field of Artificial Intelligence is a relatively new development as such this
fieldwehas borrowed discipline
field from the existing disciplines and fields.
red heavily

ARTIFICIAL INTELLIGENCE I KiPS

131
 Artificial Intelligence
Fields contributing to
Computer Science+
Computer Engineering
Mathematics+ Linguistics
Neurology + Psychology
+
Biology + Philosophy
Artificial
Intelligence
INTELLIGENCE
ARTIFICIAL
REAL LIFE USAGE OF
a number of
fields and a r e a s .
a r e being used in
in existence today
The weak Al systems
Some of these are: pre-programmed
not limiting it to
make the games m o r e interesting by experience to
Gaming: Al systems help
from their experience and
in turn, provide better gaming
systems learn
interactions. Al
the players. natural
The development in Al technology
helps in advancing the
Natural Language Processing: understand the natural
i.e., their ability to
capabilities of computer systems,
language processing (NLP)
human beings.
languages spoken by future helps
to predict or
make forecasts relating to the
Decision Making: The capability of Al systems patients.
making better decisions, for example, diagnosing
the organisations in data.
and understanding visual
in many real situations by interpreting
Vision Systems: Al systems help smart nannies for protecting
intrusions in restricted areas, identifying fugitives,
For example, identifying
children, etc. communicate
robots to
In this area, the Al systems are allowing computers and
Speech Recognition:
in a natural language
without using heavy syntax.
editable
handwritten text into computer
The Al helps in converting the
documents.This has
Handwriting Recognition: restoration of ancient text and
text.This has dramatically helped in preserving and
ARTIFICIAL INTELLIGENCE
132 Kips I
 What Al systems can do

op 1ouuD3 swajsks 1V JDuM