Artificial Intelligence

Lecture No. 3
Dr. Asad Safi

Assistant Professor,
Department of Computer Science,
COMSATS Institute of Information Technology (CIIT)
Islamabad, Pakistan.
 Summary of Previous Lecture
• AI Applications
– improvements in hardware and algorithms
– AI applications in industry, finance, medicine, and science.
• Human Intelligence VS Artificial Intelligence
– Artificial Intelligence VS Conventional Computing
• Is AI dangerous?
– Sentient AI
 Today’s Lecture
• Weak and Strong AI
• Acting humanly
• Think like humans
• think rationally
• Acting rationally
• Turing Test
• Chinese Room Argument
 What is AI?
The exciting new effort to make “The study of mental faculties
computers thinks … machine with minds, in through the use of computational
the full and literal sense” models”
(Haugeland 1985)
The automation of activities that we
(Charniak et al. 1985)
associate with human thinking, activities The study of the computations that
such as decision-making, problem solving, make it possible to perceive, reason,
learning ...'' (Bellman, 1978) and act'' (Winston, 1992)
“The art of creating machines that A field of study that seeks to explain and
perform functions that require emulate intelligent behavior in terms of
intelligence when performed by people” computational processes” (Schalkol,
(Kurzweil, 1990) 1990)
The study of how to make computers do things The branch of computer science that is
at which, at the moment, people are better'' concerned with the automation of intelligent
(Rich and Knight, 1991) behavior'' (Luger and Stubblefield, 1993)
 What is AI?
The exciting new effort to “The study of mental
make computers thinks … faculties through the use of
machine
Systems with minds,
that think inhumans
like the computational
Systems that models”
think rationally
full and literal sense” (Charniak et al. 1985)
(Haugeland 1985)

“The art of creating A field of study that seeks

machines that perform to explain and emulate
functions thatact
Systems that require
like humans intelligent
Systems behavior in terms
that act rationally
intelligence when performed of computational processes”
by people” (Kurzweil, 1990) (Schalkol, 1990)
 Weak and Strong AI
 Weak AI
 Computers can be programmed to act as if they were
intelligent (as if they were thinking)
 Strong AI
 Computers can be programmed to think (i.e. they really
are thinking)
 Weak and Strong AI
• Weak AI is AI that can not 'think', i.e. a computer chess
playing AI does not think about its next move, it is
based on the programming it was given, and its moves
depend on the moves of the human opponent.
• Strong AI is the idea/concept that we will one day
create AI that can 'think' i.e. be able to play a chess
game that is not based on the moves of the human
opponent or programming, but based on the AI's own
'thoughts' and feelings and such, which are all
supposed to be exactly like a real humans thoughts and
emotions and stuff.
 What is AI?
The exciting new effort to “The study of mental
make computers thinks … faculties through the use of
machine
Systems with minds,
that think inhumans
like the computational
Systems that models”
think rationally
full and literal sense” (Charniak et al. 1985)
(Haugeland 1985)

“The art of creating A field of study that seeks

machines that perform to explain and emulate
functions thatact
Systems that require
like humans intelligent
Systems behavior in terms
that act rationally
intelligence when performed of computational processes”
by people” (Kurzweil, 1990) (Schalkol, 1990)
 Acting humanly
The Turing Test approach

• Turing (1950) "Computing machinery and intelligence"

• The Turing Test
 Alan Turing

Alan Turing, born at 23rd of june, 1912

“I believe that in about fifty years’ time it will
be possible to programme computers, with a
storage capacity of about 109, to make them play
the imitation game so well that an average
interrogator will not have more than 70 per cent
chance of making the right identification after
5 minutes of questioning”
-Alan Turing (1950)
 Turing Test
• “Turing was convinced that if a computer could do all
mathematical operations, it could also do anything a
person can do“
• Computing Machinery and Intelligence, written
by Alan Turing and published in 1950 in Mind, is a
paper on the topic of artificial intelligence in which
the concept of what is now known as the Turing
test was introduced to a wide audience.
 The Turing Test
• Today the Game is usually referred to as the
Turing Test.
• If a computer can play the game just as well as
a human, then the computer is said to ‘pass’
the ‘test’, and shall be declared intelligent.
 Turing Test
• How can we evaluate intelligence?
– Turing [1950]: a machine can be deemed
intelligent when its responses to interrogation
by a human are indistinguishable from those of
a human being.
 Turing Test

Interrogator
 Turing Test

Interrogator
 total Turing Test
• includes a video signal so that the interrogator
can test the subject's perceptual abilities, as
well as the opportunity for the interrogator to
pass physical objects ``through the hatch.''
• To pass the total Turing Test, the computer will
need
– computer vision to perceive objects, and
– robotics to move them about.
Turing Test
 How effective is this test?
• Agent must:
– Have command of language
– Have wide range of knowledge
– Demonstrate human behavior (humor, emotion)
– Be able to reason
– Be able to learn
• Loebner prize competition is modern version of Turing Test
– (The Loebner Prize is an annual competition in artificial
intelligence that awards prizes to the chatterbot considered by
the judges to be the most human-like.)
– Example: Alice, Loebner prize winner for 2000 and 2001
 Turing Test: Criticism
• What are some potential problems with the
Turing Test?
– Some human behavior is not intelligent
• the temptation to lie, a high frequency of typing mistakes
– Some intelligent behavior may not be human
• If it were to solve a computational problem that is practically
impossible for a human to solve
– Human observers may be easy to fool
• A lot depends on expectations
• Chatbots, e.g., ELIZA, ALICE
– Chinese room argument
• Is passing the Turing test a good
scientific/engineering goal?
 Chinese Room Argument

• Devised by John Searle

• An argument against the
possibility of true
artificial intelligence.
Chinese Room Argument
 Chinese Room Argument

“The reason that no computer program can ever

be a mind is simply that a computer program is
only syntactical, and minds are more than
syntactical. Minds are semantical, they have
content.” - John Searle
 What is AI?
The exciting new effort to “The study of mental
make computers thinks … faculties through the use of
machine
Systems with minds,
that think inhumans
like the computational
Systems that models”
think rationally
full and literal sense” (Charniak et al. 1985)
(Haugeland 1985)

“The art of creating A field of study that seeks

machines that perform to explain and emulate
functions thatact
Systems that require
like humans intelligent
Systems behavior in terms
that act rationally
intelligence when performed of computational processes”
by people” (Kurzweil, 1990) (Schalkol, 1990)
 Acting humanly
The Turing Test approach
• Turing (1950) "Computing machinery and intelligence"
• The Turing Test

• What capabilities would a computer need to have to pass the Turing

Test?
– Natural language processing
– Knowledge representation
– Automated reasoning
– Machine learning
• Turing predicted that by the year 2000, machines would be able to
fool 30% of human judges for five minutes
 Thinking humanly
The cognitive modeling approach
• Goal: Develop precise theories of human
thinking
• Cognitive Architecture
– Software Architecture for modeling human performance
– Describe task, required knowledge, major sub-goals
– Architecture follows human-like reasoning
– Makes testable predictions: Time delays during problem
solving, kinds of mistakes, eye movements, verbal
protocols, learning rates, strategy shifts over time, etc.
• Problems:
– It may be impossible to identify the detailed structure of
human problem solving using only externally-available
data.
 Thinking humanly
The cognitive modelling approach
• We need to get inside the actual workings of human
minds.
• There are two ways to do this: through
• trying to catch our own thoughts as they go by
• or through psychological experiments.
• Cognitive science: the brain as an information processing
machine
– Requires scientific theories of how the brain works
• How to understand cognition as a computational process?
– try to think about how we think
– Predict and test behavior of human subjects
– Image the brain, record neurons
• The latter two methodologies are the domains of
cognitive science and cognitive neuroscience
 Thinking rationally
The laws of thought approach
• Idealized or “right” way of thinking
• Logic: patterns of argument that always yield correct conclusions
when supplied with correct premises
– “Tom is a man; all men are mortal; therefore Tom is
mortal.”
• Beginning with Aristotle, philosophers and mathematicians have
attempted to formalize the rules of logical thought
• Logicist approach to AI: describe problem in formal logical
notation and apply general deduction procedures to solve it
• Problems with the logicist approach
– Computational complexity of finding the solution
– Describing real-world problems and knowledge in logical notation
– Dealing with uncertainty
– A lot of intelligent or “rational” behavior has nothing to do with logic
 Thinking Rationally:
The Logical Approach
 Ensure that all actions performed by computer are
justifiable (“rational”)

Facts and Rules in Theorem Prover

Formal Logic

 Rational = Conclusions are provable from inputs and prior

knowledge
 Problems:
 Representation of informal knowledge is difficulty
 Hard to define “provable” reasoning
 Acting rationally
Rational agent
• A rational agent is one that acts to achieve the best
expected outcome
• Goals are application-dependent and are
expressed in terms of the utility of outcomes
• Being rational means maximizing your expected
utility
• In practice, utility optimization is subject to the
agent’s computational constraints
 Acting Rationally
Rational Agents

 Claim: “Rational” means more than just logically justified. It

also means “doing the right thing”

Rational agents do the best they can given their resources

 Weak and Strong AI
The exciting new effort to “The study of mental
make computers thinks … faculties through the use of
machine
Systems with minds,
that think inhumans
like the computational
Systems that models”
think rationally
full and literal sense” (Charniak et al. 1985)
(Haugeland 1985)
 Strong AI  Weak AI
“The art of creating A field of study that seeks
machines that perform to explain and emulate
functions thatact
Systems that require
like humans intelligent
Systems behavior in terms
that act rationally
intelligence when performed of computational processes”
by people” (Kurzweil, 1990) (Schalkol, 1990)
 Summery of Today’s Lecture
• Weak and Strong AI
• Acting humanly
• Think like humans
• think rationally
• Acting rationally
• Turing Test
• Chinese Room Argument