[Humanist] 23.597 why chess for AI
Humanist Discussion Group
willard.mccarty at mccarty.org.uk
Tue Jan 26 07:34:20 CET 2010
Humanist Discussion Group, Vol. 23, No. 597.
Centre for Computing in the Humanities, King's College London
Submit to: humanist at lists.digitalhumanities.org
Date: Mon, 25 Jan 2010 08:30:09 +0000
From: Willard McCarty <willard.mccarty at mccarty.org.uk>
Subject: why chess for AI
The fascination with chess as a problem for artificial intelligence goes
back at least as far as a paper by Claude Shannon (who famously also
provoked interest in machine translation a couple of years earlier),
"Programming a Computer for Playing Chess", Philosophical Magazine, ser.
7, 41.314 (March 1950). "Although perhaps of no practical importance,"
he wrote, "the question is of theoretical interest, and it is hoped that
a satisfactory solution of this problem will act as a wedge in attacking
other problems of a similar nature and of greater significance." He then
> (1) Machines for designing filters, equalizers, etc.
> (2)Machines for designing relay and switching circuits.
> (3) Machines which will handle routing of telephone calls based on
> circumstances rather than by fixed patterns.
> (4) Machines for performing symbolic (non-numerical) mathematical
> (5) Machines capable of translating from one language to another.
> (6) Machines for making strategic decisions in simplified military
> (7) Machines capable of orchestrating a melody.
> (8) Machines capable of logical deduction.
He points out that chess is a good place to start because,
> (1) the problem is sharply defined both in allowed operations (the
> moves) and in the ultimate goal (checkmate);
> (2) it is neither so simple as to be trivial nor too
difficult for satisfactory solution;
> (3) chess is generally considered to require "thinking" for skilful
> play; a solution of this problem will force us either to admit the
> possibility of a mechanized thinking or to further restrict our
> concept of "thinking";
> (4) the discrete structure of chess fits well
> into the digital nature of modern computers.
Feigenbaum and Feldman, Computers and Thought (1963): 4-6, also discuss
the reasons for going after chess; they emphasize how great a problem it is.
In his great lecture in Paris in 1900, "Mathematische Probleme", in
which he set out the programme for mathematics for the next generations,
David Hilbert wisely invoked the second of Shannon's principles --
neither too difficult nor too easy -- to justify his choice of problems:
"a mathematical problem should be difficult in order to entice us, yet
not completely inaccessible, lest it mock our efforts". A wise
principle, I think (which I'd do well better to respect!).
But there's another essential point, made by Pamela McCorduck in
Machines Who Think, pp. 146f: Games, she points out, are at the heart of
us and at the heart of AI. "Games are models of situations in life, just
as physical models imitate, simplify, and express the essence of
physical phenomena." Or, put another way, games simplify life by
reducing roles into rules (which approach algorithms). As children we
learn about life by playing games; I'd suppose that the rules then teach
us about the roles we observe. But I also suppose that much more has
been written along these lines, and perhaps someone familiar with the
literature would care to comment?
Willard McCarty, Professor of Humanities Computing,
King's College London, staff.cch.kcl.ac.uk/~wmccarty/;
Editor, Humanist, www.digitalhumanities.org/humanist;
Interdisciplinary Science Reviews, www.isr-journal.org.
More information about the Humanist