Artificial Intelligence:
Will robots inherit the earth?
Prabhas Chongstitvatana
Department of Computer Engineering
Chulalongkorn University
This is the first draft of the talk
AI lecture Sat 17 Nov 2001
Abstract
It is impossible to elaborate any field of knowledge as broad as Artificial
Intelligence in a few hours. What I try to do is to give a general
view of AI accompanied with some demos of the research results in the field.
My selection of topics to discuss is definitely bias, depends on my schooling.
I just don't have deep enough knowledge to qualify to talk on all topics
in AI. I shall try to address some broad claim such as "Can machines
really think the way humans do?". I hope to show practical aspect
of AI in our society.
Outline
References
What is "intelligence"
The triumph of DeepBlue over Kasparov is the public demonstration of "Machine
Intelligence" equals to human.
Does the machine "think"?
Turing Test:
A. M Turing, born 23 June 1912, London; Died 7 June 1954, Manchester
England; Pioneer in developing computer logic as we know it today. One
of the first to approach the topic of artificial intelligence. He creates
the concept of "The Turing Machine" and "Turing's Test." As
a mathematician he applied the concept of the algorithm to digital computers.
His research into the relationships between machines and nature created
the field of artificial intelligence.
<chat with Eliza>
Definition of intelligence by John McCarthy:
Intelligence is the computational part of the ability to achieve goals
in the world. Varying kinds and degrees of intelligence occur in people,
many animals and some machines.
What is AI?
David B. Leake: Indiana University. [To appear, Van Nostrand Scientific
Encyclopedia, Ninth Edition, Wiley, New York, 2002.] "Artificial intelligence
(AI) is a branch of computer science that studies the computational requirements
for tasks such as perception, reasoning, and learning, and develops systems
to perform those tasks. AI is a diverse field whose researchers address
a wide range of problems, use a variety of methods, and pursue a spectrum
of scientific goals."
John McCarthy (Stanford)
It is the science and engineering of making intelligent machines, especially
intelligent computer programs. It is related to the similar task of using
computers to understand human intelligence, but AI does not have to confine
itself to methods that are biologically observable.
What AI can do for the society
Why AI?
. . .To achieve their full impact, computer systems must have more than
processing power--they must have intelligence. They need to be able to
assimilate and use large bodies of information and collaborate with and
help people find new ways of working together effectively. The technology
must become more responsive to human needs and styles of work, and must
employ more natural means of communication.
--Barbara Grosz and Randall Davis
. . . Exactly what the computer provides is the ability not to be rigid
and unthinking but, rather, to behave conditionally. That is what it means
to apply knowledge to action: It means to let the action taken reflect
knowledge of the situation, to be sometimes this way, sometimes that, as
appropriate. . . .
In sum, technology can be controlled especially if it is saturated
with intelligence to watch over how it goes, to keep accounts, to prevent
errors, and to provide wisdom to each decision.
-Allen Newell, from Fairy Tales
AI and creativity
Harold Cohen: U. of California - San Diego, Computer
programs that create art. <toronto/aaron.html>
Excerpt from Robert MATTHEWS is science correspondent of The
Sunday Telegraph. New Scientist Volume 144. Issue 1955.
What is the creative act?
Francis Crick described '. . the dramatic feeling of the sudden
enlightenment that floods the mind when the right idea finally clinches
into place'
Margaret Boden, professor of psychology and philosophy at the
University of Sussex, rejects the idea that creativity is just the novel
combination of old ideas. If this were true, a program which simply substitutes
one formula for another logically equivalent one in a theorem could be
considered 'creative'. Boden argues that creativity involves the expansion
of a field of endeavour using ideas which could not emerge simply by following
the usual rules.
Douglas Lenat: AM (Artificial Mathematician), Eurisko
(Program that learn and invent new rules)
In the mid-1970s, a young computer scientist at Stanford University
named Douglas Lenat created a stir by unveiling AM, a program said to discover
fundamental results in mathematics. Based on the Lisp programming language,
AM had been provided with a collection of very basic concepts drawn from
mathematical set theory, such as 'intersection' and 'union', combined with
about 200 'heuristics' - or rules - for such tasks as proposing new things
to do, checking truths and spotting regularities. Once set running, AM
used these heuristics to look for new concepts that emerged, and decide
what constituted an 'interesting' discovery. AM discovered - or, more precisely,
rediscovered - the existence of addition, multiplication, de Morgan's rules
of Boolean algebra, the existence of prime numbers and the concept of unique
factorisation. Perhaps most astonishing of all, AM suggested that every
even number greater than 4 can be written as the sum of two prime numbers.
This is Goldbach's conjecture, a famous unproven idea in number theory
first put forward by a Prussian mathematician in the 18th century.
Lenat himself went on to develop Eurisko, a program which, unlike AM,
could evolve new rules for discovery as it went along. Eurisko proved particularly
effective when overseen by a human able to weed out duff ideas. Helped
by Lenat, Eurisko even won a war game, using a strategy based on small,
fast-moving attack vessels. Human competitors thought it ludicrous - until
Eurisko beat them all.
Humour
Kim Binsted: at Edinburgh University, AI researcher Kim Binsted
has developed Jape-1, a program for telling jokes. The program builds up
the jokes according to simple 'templates', such as 'What do you get if
you cross an X with a Y ?', and chooses words for X, Y and the pay-off
word Z according to properties of the words, such as their sound and associations.
Can you spot the Jape-1 jape, and the two from The Crack-a-joke Book
by human joke-merchant Kaye Webb?
A What do you give a hurt lemon? Lemonade.
B What kind of tree can you wear? A fir coat.
C What runs around a forest making other animals yawn? A wild boar.
William Chamberlain: In 1984, the New-York based writer and computer
programmer William Chamberlain published 'The Policeman's Beard is Half-constructed',
the collected works of Racter, a program that uses the basic rules of syntax
to bolt together random words and phrases. Some of Racter's output are
eerily reminiscent of the musings of French philosophers. Can you decide
which of these quotes is from Racter, and which are pearls of wisdom from
the influential French social philosopher Simone Weil?
A 'Distance is the soul of reality'
B 'Reflections are images of tarnished aspirations'
C 'Love is not consolation, it is light'
(Answers below:)
1 A is by Jape-1
2 B is by Racter
Branches of AI
According to Prof. John McCarthy: These are branches of AI
-
logical
-
search
-
pattern recognition
-
representation
-
inference
-
common senseknowledge and reasoning
-
learning from experience
-
planning
-
epistemology
-
ontology
-
heuristics
The study of AI has expanded into several subfields. These are some
of the well-known:
Ariticial Neuron Networks
<from intro/approaches.html>
The human brain is made up of a web of billions of cells called neurons,
and understanding its complexities is seen as one of the last frontiers
in scientific research. It is the aim of AI researchers who prefer this
bottom-up approach to construct electronic circuits that act as neurons
do in the human brain. Although much of the working of the brain remains
unknown, the complex network of neurons is what gives humans intelligent
characteristics. By itself, a neuron is not intelligent, but when grouped
together, neurons are able to pass electrical signals through networks.
A century earlier the true / false nature of binary numbers was theorized
in 1854 by George Boole in his postulates concerning the Laws of
Thought. Boole's principles make up what is known as Boolean algebra, the
collection of logic concerning AND, OR, NOT operands. For example according
to the Laws of thought the statement: (for this example consider all apples
red)
Apples are red-- is True
Apples are red AND oranges are purple-- is False
Apples are red OR oranges are purple-- is True
Apples are red AND oranges are NOT purple-- is also True
Boole also assumed that the human mind works according to these laws,
it performs logical operations that could be reasoned. Ninety years later,
Claude
Shannon applied Boole's principles in circuits, the blueprint for electronic
computers. Boole's contribution to the future of computing and Artificial
Intelligence was immeasurable, and his logic is the basis of neural networks.
McCulloch and Pitts, using Boole's principles, wrote a paper
on neural network theory. The thesis dealt with how the networks of connected
neurons could perform logical operations. It also stated that, one the
level of a single neuron, the release or failure to release an impulse
was the basis by which the brain makes true / false decisions. Using the
idea of feedback theory, they described the loop which existed between
the senses ---> brain ---> muscles, and likewise concluded that Memory
could be defined as the signals in a closed loop of neurons.
How a NN is trained
NetTalk is a program that used NN to learn to pronounce a word by
mapping text to phoenemes
<Rivest ML lecture "lecture16-nettalk.ps">
Artificial Life
Artificial Life relates to Biology in much the same fashion that Artificial
Intelligence relates to Psychology. The goal of artificial life, or ALife,
is to provide a different focus for researchers in biology. Rather than
emphasize an analytic approach - attempting to understand the complex phenomena
of life by breaking them down into simpler units - ALife offers a synthetic
perspective - it begins with simple rules and concepts, and combines them
to see what complex phenomena are produced. What is surprising, and what
has helped legitimize ALife as a research pursuit, is how accurately computer
models developed through this methodology have reflected our observations
of biological life. An example may help clarify what it is we mean when
we talk about combining simple rules in computer simulation to acheive
life-like results.
Game of Life
<answer the question of biological "society" such as whether the
universe "open" or "close">
<run demo game of life>
Cellular Automata, of which Life is an example, were suggested by Stanislaw
Ulam in the 1940s, and first formalized by von Neumann. Conway's
"Game of Life" was popularized in Martin Gardner's mathematical games column
in the October 1970 and February 1971 issues of Scientific American.
(Shorter notes on life are alse given in the column in each month from
October 1970 to April 1971, and well as November 1971, January 1972, and
December 1972.)
The rules for the game of life are quite simple. The game board is a
rectangular cell array, with each cell either empty or filled. At each
tick of the clock, we generate the next generation by the following rules:
if a cell is empty, fill it if 3 of its neighbors
are filled (otherwise leave it empty)
if a cell is filled,
it dies of loneliness
if it has 1 or fewer neighbors
continues to live
if it has 2 or 3 neighbors
dies of overcrowding
if it has more than 3 neighbors
Conway has demonstrated that it is possible to construct the basic building
blocks of a computer from Life using modified glider guns. See the last
chapter of Elwyn R. Berlekamp, John H. Conway, and Richard K. Guy, "Winning
Ways", Academic Press, New York, 1982, ISBN 0-120911-507.
Philosophical issue
<from comp.ai.faq>
It useful to divide AI into two classes: strong AI and weak AI. Strong
AI makes the bold claim that computers can be made to think on a level
(at least) equal to humans. Weak AI simply states that some "thinking-like"
features can be added to computers to make them more useful tools... and
this has already started to happen (witness expert systems, drive-by-wire
cars and speech recognition software). What does 'think' and 'thinking-like'
mean? That's a matter of much debate.
John Searl "Chinese room" argument
Machines will "think" differently from human.
. . . because computers lack bodies and life experiences comparable
to humans', intelligent systems will probably be inherently different from
humans.
- David L. Waltz
Future AI
Interview with Michel Negroponte (the director of MIT Media Lab)
- 13th November, 2001
"I doubt I will ever meet (in my life time) an independent, self sufficient,
thinking robot. But I suspect my grandchildren will. In the more immediate
future I think the trend will be towards smart, interactive human helpers
and appendages - for instance smart eye glasses, smart shoes, smart tooth
brushes and even smart mini-doctors that will be tiny electronic devices
that people plant on/in their bodies to give personalized health evaluations.
Imagine a toothbrush that could "see" cavities and urge you to see your
dentist ... I'd not only buy one, I need one!"
About the speaker
Prabhas Chongstitvatana is an associate professor in the department of
computer engineering, Chulalongkorn university. He graduated with
an electrical engineering bachelor degree from Kasetsart university in
1981 and earned his doctoral degree from the department of artificial intelligence
(now division of informatics) from Edinburgh university in 1992.
His research involves robotics and vision and currently evolutionary computation,
especially its application in robot programming.
homepage: www.cp.eng.chula.ac.th/faculty/pjw
email: prabhas@chula.ac.th
Handouts (for the class)
1 Basic questions, John McCarthy, <"node1-what.html">
2 Turing paper 1950
Alan M. Turing (1950). Mind 59 (Oct 1950): 433-60. {"Originally published
by Oxford University Press on behalf of MIND (the Journal of the Mind Association),
vol. LIX, no. 236, pp. 433-60, 1950.
<"turing.html">
3 Will robots inherit the earth (Minsky), <sciam.inherit.htm>
4 Chinese room argument
<whatisai/JOHN SEARLE'S CHINESE ROOM ARGUMENT>
References
Web:
AAAI www.aaai.org
McCarthy page <node5-what.html is references>
www.generation5.org <news, robotics and AI>
Textbooks:
This list is a small selection of all fine books in AI. These
are my favourites.
-
Artificial Intelligence: A Modern Approach. A textbook by Stuart Russell
and Peter Norvig. 1995. Upper Saddle River, NJ: Prentice Hall.
-
Norvig, Peter, Paradigms of Artificial Intelligence Programming: Case Studies
in Common Lisp, Morgan Kaufmann Publishers Inc., 1992.
-
Sterling, Leon, and Shapiro, Ehud, The Art of Prolog -- Advanced Programming
Techniques, 2nd edition, The MIT Press, 1994. This one is the
text with full of AI programming and to balance the view the following
is the view of AI from the opposite continent written by my class mate.
-
Dean, Thomas and Allen, James and Aloimonos, Yiannis, Artificial Intelligence:
Theory and Practice, Addison-Wesley Publishing Company, 1995.
-
Luger, G.F. and Stubblefield, W.A., Artificial Intelligence: Structures
and Strategies for Complex Problem Solving, 3rd ed. , Addison-Wesley Publishing
Company, 1997. This is a more uptodate text.
-
Nilsson, Nils J., Principles of Artificial Intelligence, Morgan Kaufmann
Publishers Inc., 1980.
-
Nilsson, Nils, Artificial Intelligence: A New Synthesis, Morgan Kaufmann
Publishers Inc., March 1998. The first one is the one I read when I
started learning about AI, the second one is the modern uptodate version.
-
Elaine Rich & Kevin Knight, "Artificial Intelligence", 2nd edition,
McGraw-Hill, New York, 1991. ISBN 0-07-052263-4. One of the more popular
introductory texts to AI, giving a very good general overview of most AI
topics. In some places the book sacrifices depth for breadth, and a few
more recent topics are neglected.
-
Patrick Henry Winston, "Artificial Intelligence", Third Edition, Addison
Wesley, Reading, MA, 1992, ISBN 0-201-53377-4. A classic early AI text.
This text is very much hands-on, with actual toy examples.
-
Boden, Margaret A. 1977. Artificial Intelligence and Natural Man. New York:Basic
Books. Although the examples are somewhat dated, this book remains a
clear discussion of the goals and controversial issues of AI. This book
used to be the standard textbook for many universities in the 80s.
-
Simon, Herbert. 1996. Sciences of the Artificial. 3rd edition. Cambridge,
MA: MIT Press. A classic book, originally published in 1969, that examines
several presuppositions of AI. Updates throughout the book take into account
advances in cognitive science and the science of design. (H. Simon died
last year 2000, he is a pioneer in AI and a Noble laureate in Economics)