INTRODUCTION TO ARTIFICIAL

INTELLIGENCE (CT017-3-1-IAI-L-2)
WEEK 1: INTRODUCTION AND BASIC CONCEPTS OF AI
LEARNING OUTCOMES
At the end of this topic, you should be able to:
Define artificial intelligence (AI)
Define human intelligence
Relate AI problems to its origins
Explain the criteria for AI success
Identified AI concepts
TOPIC AND STRUCTURE OF
PRESENTATION

AI DEFINITIONS ORIGINS OF AI AI CRITERIA OF

SUCCESS
KEY TERMS

If you have mastered this topic, you

should be able to use the following terms
correctly in your assignments and exams:
Artificial Intelligence
Human Intelligence
Turing Test
Chinese Room theory
WHAT IS ARTIFICIAL
INTELLIGENCE?
Artificial Intelligence (AI) is a branch
of computer science devoted to the
study of computer hardware and
software designed to imitate the
thinking and reasoning powers of
the human mind.
OTHER DEFINITIONS
“…the science of making machines do things that
would require intelligence if done by humans”
- Marvin Minsky

“ …the part of computer science concerned with

designing intelligent computer systems” -
E.Feigenbaum
WHAT IS INTELLIGENCE?
In the context of human intelligence:
 Ability to learn or understand from
experience
 Ability to acquire and retain knowledge
 Able to integrate experience and
knowledge to solve problems
ARTIFICIAL INTELLIGENCE VS
HUMAN INTELLIGENCE
Comparison
Human Intelligence Artificial Intelligence
Factor
Energy 2 watts for modern machine learning machine
25 watts human brain
efficiency  
   
Humans usually learn how to manage hundreds While consuming kilowatts of energy, this machine is
Universal
of different skills during life. usually designed for a few tasks.
   
 
The time needed to teach system on each and every
Multitasking Human worker work on multiple responsibilities
response is considerably high
 
 
Even the most advanced robots can hardly compete in
Decision Humans can learn decision making from
mobility with 6 years old child. And this results we have
Making experienced scenarios.
after 60 years of research and development.
 
State Brains are Analogue Computers are digital
STRONG AI
 An artificial intelligence construct that has mental
capabilities and functions that mimic the human brain.
 Also known as full AI.
 It stipulates that a computer can be programmed to be a
human mind.
 to be intelligent in every sense of the word
 to have perception, beliefs and have other cognitive states
that are normally only ascribed to humans.
Case: Turing Test by Alan Turing
WEAK AI
 A type of artificial intelligence that is limited to a specific or narrow
area.
 Also known as narrow AI.
 Rather than try to fully emulate a human mind, weak AI focuses on
developing intelligence concerned with a particular task or field of
study.
 Weak AI lacks human consciousness, although it may be able to
simulate it at times.
 Case: John Searle's Chinese room thought experiment.
STRONG AI VS WEAK AI
Strong AI Weak AI
Meaning Refers to a hypothetical machine that Simply the view that intelligent behavior can
exhibits human cognitive abilities. be modeled and used by computers to solve
complex problems.
Functions Machines with the mind of their own Systems that are programmed to performed
which can think and accomplish complex wide range of tasks within pre-defined
problems by themselves. functions.
Thinking Exhibit strong human cognitive abilities. Do not have mind of their own.
Examples Hypothetical concept which does not Alexa and Siri
exist yet in its true form
ORIGINS OF AI
AI involves various disciplines like:
 Philosophy
 Logic – IF-THEN
 Computation – Bots, Image processing,
 Psychology – Mental Health Diagnostic Expert System
 Biology – MRI, CT scan, PET scan, X-Ray, etc.
 Etc.
AI WINTER
 AI winter is a period of reduced funding and interest in AI research.
 The term was introduced in 1984 from the public debate topic in annual
meeting of American Association of Artificial Intelligence.
 The funding for AI initiatives has experienced a series of inactive and active
regime.
 Reason why the event occurred:
 Inventor unable to achieve client promises in AI initiatives which seems more complicated
to be implemented.
 Client direct their investment elsewhere due to disappointment towards AI technology
which fail to deliver promising return on investment.
PROBLEMS IN AI
• Issues like:
 Computers can only deal with syntax but can never
understand semantics
 Chinese Room argument

 Human reasoning cannot be based on a fixed calculus

 The Lucas-Penrose argument

 Humans are proved to be inconsistent

 Godel
TURING TEST
 A method to determine whether a computer can think
like a human being.
Originally called imitation game.
The test is named after Alan Turing in 1950.
 The test is used to test the machine’s ability to exhibit
human intelligence.
TURING TEST
CHINESE ROOM THEORY
 Or Chinese Room Argument was first proposed by American philosopher John
Searle in 1980.
 The test:
 Non-Chinese speaker in a room with sufficient resources (Chinese translator,
pen, paper)
 Individual outside the room sends a Chinese note into the room.
 With sufficient knowledge, the translator replied with appropriate Chinese
characters and send back the response to individual outside the room.
 This leads to individual outside the room believe that individual in the room is a
Chinese speaker.
 Hence, Searle argued that Turing Test is inadequate.
CHINESE ROOM
THEORY
The Chinese Room Argument –
John Searle
 This question in Chinese means
" What brings happiness?“
 The answer to the question is
highlighted in red. This sentence is
from Lao-Tzu's Tao Te Ching. "Be
the stream of the universe."
CHINESE ROOM THEORY
QUICK REVIEW QUESTIONS
 In your own words, define artificial intelligence

 Find out information the Lucas-Penrose argument

mentioned in the previous slide. (to be discussed
during tutorial)
CRITERIA FOR
SUCCESS
Implementin Systems that think like
g the goals humans
of AI which Systems that think rationally
are:
Systems that act like humans
Systems that act rationally
REASONING/THINKING

Ideal
Human-Based
Rationality
Reasoning- Systems that think Systems that
Based like humans. think rationally.
Behavior- Systems that act Systems that act
Based like humans. rationally.
THINKING HUMANLY
To get machine thinking like humans, the
human minds need to be understood
Therefore, if the machine’s behaviour
matches human behaviour, it can be concluded
that some of the machine’s mechanisms may
also be operating in human.
ACTING HUMANLY
 Is based on Turing test
 i.e. ability to achieve human-level performance
in all cognitive tasks shows intelligent
behaviour
 Note: Additional exercise to be discussed
during tutorial: Explain Turing test
THINKING RATIONALLY

 Related to logical thinking

 Example: We know Einstein is a man
and all men are mortal. Therefore, we
can conclude that Einstein is mortal.
ACTING RATIONALLY
 Related to making correct inferences
 Acting to achieve one’s goal, given one’s
beliefs
 Example: We saw a nail in our path. We
know that stepping on the nail will bring
pain. Therefore, we use another path.
BRIEF HISTORY OF AI
AI APPLICATIONS AND
CONCEPTS
SUMMARY

Definition Definition of
Origins of Problems in
on Artificial Human
AI AI
Intelligence Intelligence
Alan Turing Chinese
AI Criteria AI
and Turing Room
of success Applications
Test Theory
QUESTIONS AND ANSWERS
SESSION
NEXT TOPIC

 Introduction to Expert System