Artificial ntelligence

$
Instructor: Professor Yun Peng
IET BuiIding Room 341
(410)455-3816
ypeng@umbc.edu
Some material adopted Irom notes by
Charles R. Dyer, University oI Wisconsin-Madison and
1im Finin and Marie des1argins, University oI Maryland
Baltimore County
Introduction Introduction
Chapters 1& 2
Big questions
Can machines think?
And iI so, how?
And iI not, why not?
And what does this say about human beings?
And what does this say about the mind?
,t is ,rtifici,I inteIIigence?
There are no clear consensus on the deIinition oI AI
Here`s one Irom John McCarthy, (He coined the phrase AI in
1956) - see http:// www. Iormal. StanIord. EDU/ jmc/ whatisai/)
Q. What is artificial intelligence?
A. It is the science and engineering oI making intelligent machines,
especially intelligent computer programs. It is related to the
similar task oI using computers to understand human intelligence,
but AI does not have to conIine itselI to methods that are
biologically observable.
Q. Yes, but what is intelligence?
A. Intelligence is the computational part oI the ability to achieve
goals in the world. Varying kinds and degrees oI intelligence
occur in people, many animals and some machines.
ter possibIe AI definitions
AI is a collection oI hard problems which can be solved by
humans and other living things, but Ior which we don`t have
good algorithms Ior solving.
e. g., understanding spoken natural language, medical
diagnosis, circuit design, learning, selI-adaptation,
reasoning, chess playing, proving math theories, etc.
DeIinition Irom R & N book: a program that
Acts like human (Turing test)
Thinks like human (human-like patterns oI thinking steps)
Acts or thinks rationally (logically, correctly)
Some problems used to be thought oI as AI but are now
considered not
e. g., compiling Fortran in 1955, symbolic mathematics in
1965, pattern recognition in 1970
,t's e,sy ,nd w,t's ,rd?
It`s been easier to mechanize many oI the high level cognitive
tasks we usually associate with 'intelligence in people
e. g., symbolic integration, proving theorems, playing chess,
some aspect oI medical diagnosis, etc.
It`s been very hard to mechanize tasks that animals can do easily
walking around without running into things
catching prey and avoiding predators
interpreting complex sensory inIormation (visual, aural, .)
modeling the internal states oI other animals Irom their
behavior
working as a team (ants, bees)
Is there a Iundamental diIIerence between the two categories?
Why some complex problems (e.g., solving diIIerential equations,
database operations) are not subjects oI AI
Foundations oI AI
Computer
Science &
Engineering
AI
Mathematics
Cognitive
Science
Philosophy
Psychology Linguistics
Biology
Economics
istory of AI
AI has roots in a number oI scientiIic disciplines
computer science and engineering (hardware and soItware)
philosophy (rules oI reasoning)
mathematics (logic, algorithms, optimization)
cognitive science and psychology (modeling high level
human/animal thinking)
neural science (model low level human/animal brain activity)
linguistics
The birth oI AI (1943 1956)
Pitts and McCulloch (1943): simpliIied mathematical model oI
neurons (resting/Iiring states) can realize all propositional logic
primitives (can compute all Turing computable Iunctions)
Allen Turing: Turing machine and Turing test (1950)
Claude Shannon: inIormation theory; possibility oI chess playing
computers
Tracing back to Boole, Aristotle, Euclid (logics, syllogisms)
Early enthusiasm (1952 1969)
1956 Dartmouth conIerence
John McCarthy (Lisp);
Marvin Minsky (Iirst neural network machine);
Alan Newell and Herbert Simon (GPS);
Emphasize on intelligent general problem solving
GSP (means-ends analysis);
Lisp (AI programming language);
Resolution by John Robinson (basis Ior automatic theorem
proving);
heuristic search (A*, AO*, game tree search)
Emphasis on knowledge (1966 1974)
domain speciIic knowledge is the key to overcome existing
diIIiculties
knowledge representation (KR) paradigms
declarative vs. procedural representation
Knowledge-based systems (1969 1999)
DENDRAL: the Iirst knowledge intensive system (determining
3D structures oI complex chemical compounds)
MYCIN: Iirst rule-based expert system (containing 450 rules Ior
diagnosing blood inIectious diseases)
EMYCIN: an ES shell
PROSPECTOR: Iirst knowledge-based system that made
signiIicant proIit (geological ES Ior mineral deposits)
AI became an industry (1980 1989)
wide applications in various domains
commercially available tools
Current trends (1990 present)
more realistic goals
more practical (application oriented)
distributed AI and intelligent soItware agents
resurgence oI neural networks and emergence oI genetic
algorithms
Possible Approaches
Think
Act
Like
humans
Well
GPS
Eliza
Rational
agents
Heuristic
systems
AI tends to
work mostly
in this area
%hink well
Develop Iormal models oI knowledge
representation, reasoning, learning,
memory, problem solving, that can be
rendered in algorithms.
There is oIten an emphasis on a systems that are
provably correct, and guarantee Iinding an optimal
solution.
Think
Act
Like
humans Well
GPS
Eliza
Rational
agents
Heuristic
systems
Act well
For a given set oI inputs, generate an
appropriate output that is not necessarily
correct but gets the job done.
A heuristic (heuristic rule, heuristic method) is a rule oI
thumb, strategy, trick, simpliIication, or any other kind oI
device which drastically limits search Ior solutions in large
problem spaces.
Heuristics do not guarantee optimal solutions; in Iact, they
do not guarantee any solution at all: all that can be said for
a useful heuristic is that it offers solutions which are
good enough most of the time.
- Feigenbaum and Feldman, 1963, p. 6
Think
Act
Like
humans Well
GPS
Eliza
Rational
agents
Heuristic
systems
%hink like humans
Cognitive science approach
Focus not just on behavior and I/O
but also look at reasoning process.
Computational model should reIlect 'how results were
obtained.
Provide a new language Ior expressing cognitive theories
and new mechanisms Ior evaluating them
GPS (General Problem Solver): Goal not just to produce
humanlike behavior (like ELIZA), but to produce a
sequence oI steps oI the reasoning process that was similar
to the steps Iollowed by a person in solving the same task.
Think
Act
Like
humans Well
GPS
Eliza
Rational
agents
Heuristic
systems
Act like humans
Behaviorist approach.
Not interested in how you get results, just the similarity to
what human results are.
ExempliIied by the Turing Test (Alan Turing, 1950).
Think
Act
Like
humans Well
GPS
Eliza
Rational
agents
Heuristic
systems
%uring %est
Three rooms contain a person, a computer, and an
interrogator.
The interrogator can communicate with the other two by
teleprinter.
The interrogator tries to determine which is the person and
which is the machine.
The machine tries to Iool the interrogator into believing that
it is the person.
II the machine succeeds, then we conclude that the machine
can think.
Eliza
ELIZA: A program that simulated a psychotherapist
interacting with a patient and successIully passed the Turing
Test.
Coded at MIT during 1964-1966 by Joel Weizenbaum.
First script was DOCTOR.
The script was a simple collection oI syntactic patterns not unlike
regular expressions
Each pattern had an associated reply which might include bits oI the
input (aIter simple transIormations (my Fyour)
Weizenbaum was shocked at reactions:
Psychiatrists thought it had potential.
People unequivocally anthropomorphized.
Many thought it solved the NL problem.
Eliza
I am the psychotherapist. Please, describe your problems. Each
time you are finished talking, type RE% twice.
i'm having a hard time with my students
Why do you say that?
well, they keep asking me for money
Maybe your plans have something to do with this.
what plans?
Is it because of your life that you say what plans?
No, it's because I didn't know what you were talking about
Is it because you didn't know what I was talking about that you
came to me?
no, it's because i wanted to see if you could pass the turing test
Does it bother you that you wanted to see if I could really pass
the turing test?
...
How do you reconcile problems at school?
%his is a waste of time. Good bye
My secretary will send you a bill.
I am the psychotherapist. Please, describe your problems. Each
time you are finished talking, type RE% twice.
i'm having a hard time with my students
Why do you say that?
well, they keep asking me for money
Maybe your plans have something to do with this.
what plans?
Is it because of your life that you say what plans?
No, it's because I didn't know what you were talking about
Is it because you didn't know what I was talking about that you
came to me?
no, it's because i wanted to see if you could pass the turing test
Does it bother you that you wanted to see if I could really pass
the turing test?
...
How do you reconcile problems at school?
%his is a waste of time. Good bye
My secretary will send you a bill.
This is a
transcript oI a
conversation
with an Eliza-
like program
Iound in
Emacs
Try typing
Meta-x
doctor
in your Iavorite
Emacs.
Colby`s PARRY
Kenneth Colby modeled
a paranoid using the
same techniques circa
1968.
PARRY has basic
emotions. II it gets
angry, its replies become
more hostile.
In a Turing test, expert
psychiatrists were unable
to distinguish PARRY`s
ramblings Irom those oI
real paranoids.
H: Good aIternoon. My name is
Doctor Jones.
P: Pleased to meet you.
H: How are you?
P: All right I guess.
H: Could you tell me your name?
P: Pat Smith.
H: I though you might have some
problems.
P: I've been upset lately.
H: By what?
P: People get on my nerves
sometimes.
H: Tell me more.
P: Do you know anything about
bookies?
...
H: Good aIternoon. My name is
Doctor Jones.
P: Pleased to meet you.
H: How are you?
P: All right I guess.
H: Could you tell me your name?
P: Pat Smith.
H: I though you might have some
problems.
P: I've been upset lately.
H: By what?
P: People get on my nerves
sometimes.
H: Tell me more.
P: Do you know anything about
bookies?
...
Parry
meets
Eliza
P: I don't understand your motives.
E: Why don't you understand my motives?
P: Shouldn't I be?
E: Why do you say that just now?
P: That's just the way it is.
E: Suppose that's the way it were not.
P: So?
E: I am not sure I understand you Iully.
P: You harp on the same questions too much.
E: What does that similarity suggest to you?
P: I don't get you.
E: Do you wish that you did get I?
P: I don't understand your motives.
E: Why don't you understand my motives?
P: Shouldn't I be?
E: Why do you say that just now?
P: That's just the way it is.
E: Suppose that's the way it were not.
P: So?
E: I am not sure I understand you Iully.
P: You harp on the same questions too much.
E: What does that similarity suggest to you?
P: I don't get you.
E: Do you wish that you did get I?
%he Loebner Contest
A modern version oI the Turing Test, held annually, with a
$100,000 cash prize.
Hugh Loebner was once director oI UMBC`s Academic
Computing Services (nee UCS)
http://www.loebner.net/PrizeI/loebner-prize.html
Restricted topic (removed in 1995) and limited time.
Participants include a set oI humans and a set oI computers
and a set oI judges.
Scoring
Rank Irom least human to most human.
Highest median rank wins $2000.
II better than a human, win $100,000. (Nobody yet.)
What can AI systems do
Here are some example applications
Computer vision: Iace recognition Irom a large set
Robotics: autonomous (mostly) automobile
Natural language processing: simple machine translation
Expert systems: medical diagnosis in a narrow domain
Spoken language systems: ~1000 word continuous speech
Planning and scheduling: Hubble Telescope experiments
Learning: text categorization into ~1000 topics
User modeling: Bayesian reasoning in Windows help (the
inIamous paper clip.)
Games: Grand Master level in chess (world champion),
checkers, etc.
What can`t AI systems do yet?
Understand natural language robustly (e.g., read and
understand articles in a newspaper)
SurI the web
Interpret an arbitrary visual scene
Learn a natural language
Play Go well
Construct plans in dynamic real-time domains
ReIocus attention in complex environments
PerIorm liIe-long learning