TAI2151 Artificial Intelligence

Fundamentals

LECTURE 1: INTRODUCTION
 Text book and reference books
Text Book

 Stuart Russell, Peter Norvig (2021). Artificial Intelligence: A Modern Approach (4th ed.). Pearson.
Reference books
 1. Danny Kopec, Shweta Shetty, Christopher Pileggi (2014). Artificial Intelligence Problems and Their
Solutions (1st ed.). Mercury Learning & Information.
 2. Kevin Warwick (2011). Artificial Intelligence: The Basics. Routledge.
 3. David L. Poole, Alan K. Mackworth, (2010). Artificial Intelligence: Foundations of Computational
Agents. Cambridge University Press.
 4. George Luger (2008). Artificial Intelligence: Structures and Strategies for Complex Problem Solving (6th
ed.). Addison-Wesley.
 5. M. Tim Jones (2008). Artificial Intelligence: A Systems Approach (2008). Jones and Bartlett Publishers.
 6. Ivan Bratko (2011). Prolog Programming for Artificial Intelligence (4th ed.). Pearson. 7. David S.
Touretzky (2013). Common LISP: A Gentle Introduction to Symbolic Computation (Revised Edition). Dover
Publications. 8. Paul Graham (2001). ANSI Common LISP (2nd ed.). Prentice Hall.
What are we striving for?

The objectives of artificial intelligence are

 To model and process intelligence so as to build intelligent
entities (systems).
 Scientific motivations for creation of new exact algorithms
and heuristics.
Different faces of AI
 Symbolic Computation,
 Computational Intelligence: Evolutionary
computation, neural networks, fuzzy logic etc.
 Emphasis in the course will be the first approach with some
mention of other approaches
What is AI ?

 Thinking like humans: decision-making, problem

solving, learning
 Thinking rationally: study of mental faculties through
computational models
 Acting like humans: perform functions requiring
intelligence of humans
 Acting rationally: automation of intelligent behaviour.
 In short, we try to use computational techniques to
perform tasks that required the intelligence of human.
For e.g.: making decision to move forward or backward
What is Intelligence?
 Use of intuition (to sense what a solution must be
right just because it feels right)
 Use of common sense (logic)
 Use of judgement
 Use of creativity
 Use of reasoning
 Use of knowledge
 Human obtain knowledge by experience and learning
and demonstrate their intelligence to solve problems
Rationality

 A system is rational if it does the right thing

 A system that can understand human language
 A system that can recognize a face
AI ?
Approaches to AI
 Search
 Learning
 Rule-Based Systems
 Reasoning (logic)
 Planning
 Ability-Based Areas
 Robotics
 Agent
Search

 “All AI is search” – search to get the goal

 Game theory
 Problem spaces

 Every problem is a “virtual” tree of all possible

(successful or unsuccessful) solutions.
 The trick is to find an efficient search strategy.
Search: Game Theory

9!+1 = 362,880
Approaches to AI
 Search
 Learning
 Rule-Based Systems
 Reasoning (logic)
 Planning
 Ability-Based Areas
 Robotics
 Agent
Learning

 Learn from the available data (experience, past history)

 Learning that is explainable:
 Discovery
 Data Mining
 Learning that is non-explainable:
 Neural Nets (black box)
 Case-Based Reasoning
 Develop solution based on pre-existing solution of similar nature
Learning: No Explanation

 Neural nets
Approaches to AI
 Search
 Learning
 Rule-Based Systems
 Reasoning (logic)
 Planning
 Ability-Based Areas
 Robotics
 Agent
Rule-Based Systems

 Logic Languages
 Prolog, Lisp
 Knowledge bases
 Inference engines
Rule-Based Languages: Prolog

Father(abraham, isaac). Male(isaac).

Father(haran, lot). Male(lot).
Father(haran, milcah). Female(milcah).
Father(haran, yiscah). Female(yiscah).
Son(X,Y)  Father(Y,X), Male(X).
Daughter(X,Y)  Father(Y,X), Female(X).

Son(lot, haran)?
Approaches to AI
 Search
 Learning
 Rule-Based Systems
 Reasoning Logic
 Planning
 Ability-Based Areas
 Robotics
 Agents
Ability-Based Areas
 Computer vision
 Natural language recognition
 Natural language generation
 Speech recognition
 Speech generation
 Robotics
 Agents
Natural Language: Translation

“Behavior is primarily adaptation to the

environment under sensory guidance. It takes
the organism away from harmful events and
towards favourable ones, or introduces
changes in the immediate environment that
make survival more likely. ” Hebb
“The food was lousy, but the drink was
great!”
 Translate to Malay
 Translate back to English
Natural Language Recognition
OBJ GOLD: X

PERSON: PERSON:
Semantics REPT TRANSACTION AGNT Fred
Joe

Context

sentence
w

VP
VP
NP

Syntax VP NP NP

pronoun verb pronoun article noun

n d

Words You give me the gold

Audio
 Natural Language Recognition

PERSON: BELIEF
“Tom Tom
EXPR

believes
Mary PTNT

wants to
marry a PROPOSITION
sailor.” :
PERSON:
EXPR WANT
Mary

PTNT

SITUATION:

T AGNT MARRY PTNT SAILOR

Turing Test
 The Turing test is a behavioural approach to determining
whether or not a system is intelligent.
 Curious to know: if human use available information to solve
problem and make decision, can machine doe the same as
well?
Turing Test Continued ...

Modified Turing Test Setup

Intelligence Continued ...
 Turing Test: A human communicates with a computer via a
teletype. If the human can’t tell he is talking to a computer
or another human, it passes.
 Natural language processing
 knowledge representation
 automated reasoning
 machine learning
 Add vision and robotics to get the total Turing test.
 Try Alice the Chatterbot.
Weak and Strong AI Claims

 Weak AI:
 Machines can be made to act as if they were intelligent.
 Strong AI:
 Machines that act intelligently have real, conscious minds.
The Chinese Room (CR) Experiment

CR argument was first published by John Searle in 1980

This experiment is to show that it is impossible for computers to understand

human language or think (Strong AI)
The Chinese Room Experiment continued...

 The Chinese Room

`
CR argument: A popular argument that has no conclusion. Some researchers agree
with Searle but some researchers do not agree
What is Intelligence?

 Replacing the brain

Intelligent Systems in Everyday
Life
 Post office – automatic address recognition
 Banks – automated cheque readers, signature
verification, fraud detection, loan application
classification.
 Telecommunications – voice recognitions for directory
enquiries, fraud detection.
 Credit card companies – fraud detection, application
screening.
 Computer companies – automated help desk diagnosis.
 Household appliances – Washing machines, rice
cookers.
The Foundation of Artificial Intelligence

 Philosophy
 Psychology
 Mathematics
 Computer Engineering
 Linguistics
The History of Artificial Intelligence
Continued ...
 Modern Founders of AI
 Alan Turing ("Computing Machinery and Intelligence";
Turing test) (1950)
 McCulloch & Pitts (Neural nets) (1943)
 Norbert Wiener (Cybernetics)
 John von Neumann (Game theory)
 Claude Shannon (Information theory)
 Newell & Simon (The Logic Theorist)
 John McCarthy (LISP, commonsense reasoning)
 Marvin Minsky (Frames)
 The History of Artificial Intelligence Continued ...

 Achievements of AI
 Deep Thought is an international grand master chess player.
 Sphinx can recognise continuous speech without training for each
speaker. It operates in near real time using a vocabulary of 1000 words
and has 94% word accuracy.
 Navlab is a car that has driven across the United States at 55mph in
normal traffic on freeways.
 Carlton and United Breweries use an AI planning system to plan
production of their beer.
 Robots are used regularly in manufacturing.
 Natural language interfaces to databases can be obtained on a PC.
 Machine Learning methods have been used to build expert systems.
 Expert systems are used regularly in finance, medicine, manufacturing
and agriculture
The History of Artificial Intelligence Continued ...
 1943-1956 (the age of invention)
 McCulloch & Pitts/Hebb (a model of brain)
 Simple neural models of processing

 Claude Shannon/Turing
 computers can manipulate symbols
 chess as a canonical example
 Dartmouth workshop (1956)
 genesis of the term “AI”
 Addressing human creativity, self learn, self improvement of
human
 Try to build machines that function automatically in complex
and changing environement
The History of Artificial Intelligence Continued ...

 Newell &Simon
 Logic Theorist

 Reasoning program
 1952-1969 (Early AI Programs)
 General Problem Solver
 E.g. Question answering system
 Samuel checkers program
 LISP by McCarthy (1958) – use knowledge to search solution

 Resolution method (1965) ~ theorem proving algorithm for first

order logic
 Source: https://sitn.hms.harvard.edu/flash/2017/history-artificial-intelligence/