Artificial Intelligence
1
AI in Fiction
An intelligent killing robot
Smart machines that took over
the human race and made
them live in a simulated world
2
What’s interesting with AI
Search engines
Science
Medicine/
Appliances Diagnosis
Labor
Movies Recommendation
3
What’s interesting with AI
• Honda AISMO
• Advanced Step in Innovation MObility
• Humanoid Robot
• Capable of recognizing:
• Moving objects
• Postures
• Gestures
• Handshake
• Sounds
• Capable of walking and running
4
What’s interesting with AI
• 1996, Deep Blue first machine to beat chess world champion
• But lost in the series – 4 to 2
• 1997, won the series 3.5 to 2.5
• Search 6 to 8 moves a head
• The evaluation function is set by the system after examining thousands of master
games
5
AI Definition
• The exciting new effort to make computers thinks …
machine with minds, in the full and literal sense”
(Haugeland 1985)
• The automation of activities that we associate with human
thinking, activities such as decision-making, problem
solving, learning,…(Bellman, 1978)
Think Like Humans
6
AI Defintion
• “The art of creating machines that perform functions that
require intelligence when performed by people” (Kurzweil,
1990)
• “The study of how to make computers do things at which, at
the moment, people do better”, (Rich and Knight, 1991)
Act Like Humans
7
AI Definition
• “The study of mental faculties through the use of
computational models”,(Charniak et al. 1985)
• “The study of the computations that make it possible to
perceive, reason and act”,(Winston, 1992)
Think Rationally
8
AI Definition
• “Computational Intelligence is the study of the design of
intelligent agents” (Poole et al, 1998)
• “AI….is concerned with intelligent behavior in artifact”,
(Nilsson, 1998)
Act Rationally
9
How to Achieve AI?
Acting
humanly
Thinking
humanly AI Thinking
rationally
Acting
rationally
10
Acting Humanly: The Turing Test
http://en.wikipedia.org/wiki/Turing_test
Alan Turing
1912-1954
• To be intelligent, a program should simply act like a human
11
The Turing Test - Example
http://aimovie.warnerbros.com http://www.ai.mit.edu/projects/infolab/
slide mostly borrowed from Laurent Itti 12
The Turing Test - Example
http://aimovie.warnerbros.com http://www.ai.mit.edu/projects/infolab/
slide mostly borrowed from Laurent Itti 13
The Turing Test - Example
http://aimovie.warnerbros.com http://www.ai.mit.edu/projects/infolab/
slide mostly borrowed from Laurent Itti 14
The Turing Test - Example
15
The Turing Test - Example
http://aimovie.warnerbros.com http://www.ai.mit.edu/projects/infolab/
slide mostly borrowed from Laurent Itti 16
Acting Humanly
• To pass the Turing test, the computer/robot needs:
– Natural language processing to communicate successfully.
– Knowledge representation to store what it knows or hears.
– Automated reasoning to answer questions and draw conclusions using
stored information.
– Machine learning to adapt to new circumstances and to detect and
extrapolate patterns.
– These are the main branches of AI.
17
Acting Humanly: The Turing Test
http://en.wikipedia.org/wiki/Turing_test
+ physical interaction =>
Total Turing Test
- Recognize objects and
gestures
- Move objects
Alan Turing
1912-1954
• To be intelligent, a program should simply act like a human
18
Acting Humanly – for Total Turing
• To pass the Turing test, the computer/robot needs:
– Natural language processing to communicate successfully.
– Knowledge representation to store what it knows or hears.
– Automated reasoning to answer questions and draw conclusions using stored
information.
– Machine learning to adapt to new circumstances and to detect and extrapolate
patterns.
– Computer vision to perceive objects. (Total Turing test)
– Robotics to manipulate objects and move. (Total Turing test)
– These are the main branches of AI.
19
Thinking Humanly
• Real intelligence requires thinking think like a
human !
• First, we should know how a human think
– Introspect ones thoughts
– Physiological experiment to understand how someone
thinks
– Brain imaging – MRI…
• Then, we can build programs and models that
think like humans
– Resulted in the field of cognitive science: a merger
between AI and psychology.
20
Problems with Imitating Humans
• The human thinking process is difficult to
understand: how does the mind raises from
the brain ? Think also about unconscious tasks
such as vision and speech understanding.
• Humans are not perfect ! We make a lot of
systemic mistakes:
21
Thinking Rationally
• Instead of thinking like a human : think rationally.
• Find out how correct thinking must proceed: the laws
of thought.
• Aristotle syllogism: “Socrates is a man; all men are
mortal, therefore Socrates is mortal.”
• This initiated logic: a traditional and important branch
of mathematics and computer science.
• Problem: it is not always possible to model thought as
a set of rules; sometimes there uncertainty.
• Even when a modeling is available, the complexity of
the problem may be too large to allow for a solution.
22
Acting Rationally
• Rational agent: acts as to achieve the best outcome
• Logical thinking is only one aspect of appropriate behavior:
reactions like getting your hand out of a hot place is not the
result of a careful deliberation, yet it is clearly rational.
• Sometimes there is no correct way to do, yet something
must be done.
• Instead of insisting on how the program should think, we
insist on how the program should act: we care only about
the final result.
• Advantages:
– more general than “thinking rationally” and more
– Mathematically principled; proven to achieve rationality unlike
human behavior or thought
23
Acting Rationally
This is how birds fly Humans tried to mimic This is how we finally
birds for centuries achieved “artificial flight”
24
Relations to Other Fields
• Philosophy
– Logic, methods of reasoning and rationality.
• Mathematics
– Formal representation and proof, algorithms, computation, (un)decidability, (in)tractability,
probability.
• Economics
– utility, decision theory (decide under uncertainty)
• Neuroscience
– neurons as information processing units.
• Psychology/Cognitive Science
– how do people behave, perceive, process information, represent knowledge.
• Computer engineering
– building fast computers
• Control theory
– design systems that maximize an objective function over time
• Linguistics
– knowledge representation, grammar
slide mostly borrowed from Max Welling
25
AI History
• Gestation of AI (1934 - 1955)
– In 1943, proposed a binary-based model of neurons
– Any computable function can be modeled by a set of neurons
– A serious attempt to model brain
– 1950, Turing’s “Computing Machinery and Intelligence ”: turing test,
reinforcement learning and machine learning
• The Inception of AI (1956)
– Dartmouth meeting to study AI
– an AI program ”Logic Theorist” to prove many theorems
• Early Enthusiasm and great Expectation (1952-1969)
– General Problem Solver imitates the human way of thinking
– LISP (AI programming language) was defined
– 1965, Robinson discovered the resolution method – logical reasoning
• AI Winter (1966-1973)
– Computational intractability of many AI problems
– Neural Network starts to disappear
26
AI History
• Knowledge-based systems (1969-1979)
– Use domain knowledge to allow for stronger reasoning
• Becomes an Industry (1980-now)
– Digital Equipment Corporation selling R1 “expert sytem”
– From few million to billions in 8 years
• The return of neural network (1986-now)
– With the back-propagation algorithm
• AI adopts scientific method (1987-now)
– More common to base theorems on pervious ones or rigorous evidence rather
than intuition
– Speech recognition and HMM
• Emergence of intelligent agent (1995-now)
– search engines, recommender systems,….
• Availability of very large data sets (2001 – now)
– Worry more about the data
27
The State of the Art
• Robotics Vehicle
– DARPA Challenge
• Speech Recognition
– United Airlines
• Autonomous Planning and Scheduling
– Remote Agent: Plan and control spacecraft
– MAPGEN: daily planning of operations on NASA’s exploration Rover
• Game Playing
– IBM Deep Blue
• Spam Fighting
• Logistic Planning
– DART – Dynamic Analysis and Replacing Tool
– Gulf War 1991
– To plan the logistic for transportation of 50k vehicles, cargo and people
– Generated in hour a plan that could take weeks
• Robotics
• Machine Translation
– Statistical models
28
Summary
• This course is concerned with creating rational agents:
artificial rationality.
• AI has passed the era of infancy and is now attacking real
life, complex problems, and it is succeeding in many of
them.
• The history of AI has had a turbulent history with many ups
and downs, phenomenal successes and deep
disappointments resulting in fund cutbacks and economic
losses.
• AI has flourished in the last two decades and it the
researchers mentality shifted towards a rigorous scientific
methodology:
Firm theoretical basis & Serious experiments
29
