[Humanist] 23.600 why chess for AI

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Wed Jan 27 09:26:46 CET 2010


                 Humanist Discussion Group, Vol. 23, No. 600.
         Centre for Computing in the Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

  [1]   From:    James Rovira <jamesrovira at gmail.com>                       (7)
        Subject: Re: [Humanist] 23.597 why chess for AI

  [2]   From:    Amanda Gailey <amanda.gailey at gmail.com>                  (131)
        Subject: Re: [Humanist] 23.597 why chess for AI

  [3]   From:    Wendell Piez <wapiez at mulberrytech.com>                    (61)
        Subject: Re: [Humanist] 23.597 why chess for AI


--[1]------------------------------------------------------------------------
        Date: Tue, 26 Jan 2010 02:10:30 -0500
        From: James Rovira <jamesrovira at gmail.com>
        Subject: Re: [Humanist] 23.597 why chess for AI
        In-Reply-To: <20100126063420.43BAC49297 at woodward.joyent.us>

There's a story, perhaps apocryphal, about the invention of a
chess-playing "mechanical man" who was pitted against Napoleon.
Napoleon exposed the robot as a fraud by repeatedly making illegal
moves until the midget hiding inside got angry and swept the pieces
off of the board.  At any rate, game playing as a problem for AI may
predate the 20thC.  It may predate the 19thC.

Jim R



--[2]------------------------------------------------------------------------
        Date: Tue, 26 Jan 2010 11:40:15 -0600
        From: Amanda Gailey <amanda.gailey at gmail.com>
        Subject: Re: [Humanist] 23.597 why chess for AI
        In-Reply-To: <20100126063420.43BAC49297 at woodward.joyent.us>


Hi, Willard--

I'd like to point out that chess as a problem for artificial intelligence
has captured the imagination long before artificial intelligence had any
hope of tackling chess, as evidenced by "The Turk" (the 18th-century
automaton hoax in which a concealed man operated a contraption in such a way
that a mechanical "Turk" outside the box appeared to play chess).

One interesting aspect of this is how, to 18th- and 19th-century western
audiences at least, artificial intelligence seemed bound up with orientalism
and the occult: creators of automata were more interested in developing an
atmosphere of mystique and mysticism (culminating, maybe, in the
fortune-telling machines in Depression-era penny arcades?) than in calling
attention to the cleverness of the technology.  Perhaps one reason chess
lent itself to early AI was that unlike tic-tac-toe or checkers (more
suitable subjects from a practical standpoint), chess, with its eastern
origins and vaguely anthropomorphic pieces, seemed more suited to the
aesthetic.

In 1836 Edgar Allen Poe--so perfectly combining the occult with
ratiocination himself--wrote an editorial for The Southern Literary
Messenger in which he considers the Turk alongside Babbage's difference
engine. (Full article available here:
http://www.eapoe.org/works/essays/maelzel.htm)  Ultimately, Poe claims, if
the Turk were indeed fully automated, it would be a far more impressive
machine than Babbage's: Babbage's would use the simple logic of a machine to
proceed unerringly from raw data to a single inevitable output, while with
the Turk "there is no determinate progression. No one move in chess
necessarily follows upon any one other."  Poe concludes that if the Turk
proves to be "pure machine," it would, unlike Babbage's, have a mind.

This is, of course, exactly what Kasparov regrets: modern chess programs are
powerful but ultimately not much closer to truly mimicking the mind than
when Poe reflected on the problem 174 years ago.

Thanks,

Amanda

-- 
Amanda Gailey
Assistant Professor
Department of English
Center for Digital Research in the Humanities
University of Nebraska
202 Andrews Hall
Lincoln, NE 68588


--[3]------------------------------------------------------------------------
        Date: Tue, 26 Jan 2010 12:52:25 -0500
        From: Wendell Piez <wapiez at mulberrytech.com>
        Subject: Re: [Humanist] 23.597 why chess for AI
        In-Reply-To: <20100126063420.43BAC49297 at woodward.joyent.us>

Dear Willard,

At 01:34 AM 1/26/2010, you wrote:
>[Shannon] 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.

I think (3) here is very interesting. It's especially interesting 
that he suggests the alternative to "admit[ting] the possibility of 
machine thinking" (presumably because we wish thinking, by 
definition, to be something machines can't do) is "to further 
restrict our concept of thinking".

That is, either the chess-playing machine is a member of the set of 
thinkers, or chess playing is not sufficient to qualify one as a 
member of the set.

I might be tempted to choose the second option, but I'm not sure I'd 
describe this as a "restriction" of my concept of thinking. 
"Refinement" might be more like it.

Yet on the other hand, I think the point is important, particularly 
in a world in which entities other than individual people, such as 
cats or dogs, or basketball teams, bureaucracies, stock markets, 
voting publics, or entire genomes (which "learn" from one generation 
to the next) seem also to be "thinking". I would sooner say that all 
of these are thinking than I would that computers are thinking. 
Thinking is certainly involved in what the computers are doing. But 
I'm not sure the computer is quite as autonomous as we pretend, even 
after we push the button and it starts playing chess: I think its 
thinking is an extension of and largely limited by its programmers' 
and user's thinking.

>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.

Indeed. I wrote about this, in the context of markup languages, last 
year. It will be more interesting to those concerned with the 
practice and politics of markup technologies than to others, but it 
also speaks more generally about games and technologies:

http://www.balisage.net/Proceedings/vol3/html/Piez01/BalisageVol3-Piez01.html

(Check out the slides, too. They're fun.)

Cheers,
Wendell

=========================================================
Wendell Piez                            mailto:wapiez at mulberrytech.com
Mulberry Technologies, Inc.                http://www.mulberrytech.com
17 West Jefferson Street                    Direct Phone: 301/315-9635
Suite 207                                          Phone: 301/315-9631
Rockville, MD  20850                                 Fax: 301/315-8285
----------------------------------------------------------------------
   Mulberry Technologies: A Consultancy Specializing in SGML and XML
=========================================================





More information about the Humanist mailing list