Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Shredder 8 in Argentina: Seconds or Centaurs?

Author: Peter Berger

Date: 17:39:39 07/23/04

Go up one level in this thread


On July 23, 2004 at 12:25:55, George Sobala wrote:

>On July 23, 2004 at 01:50:12, Sandro Necchi wrote:
>
>>
>>Still I understant this is all the owners can do. I am not criticizing this, but
>>want to remember that there is available a stronger version, so the gap, if any,
>>is lower.
>>
>>
>>Sandro
>
>I do realise that most major computer tournaments such as WCCC and man-machine
>matches are played this way, with a programmer or team fine-tuning the computer,
>especially the book, from game to game.
>
>But to me it is an interesting philosophical question: is the human team behind
>the engine, which chooses which openings to play, merely like the seconds who
>help top players through e.g. World Championship matches, or are they more than
>this: does this team become more like a centaur, a computer-human fusion? I
>would argue that it is really a type of centaur. Kramnik's second will not
>*tell* him which openings to play against Leko, he will merely assist him with
>his preparation. Whereas in a tournament you *tell* Shredder which openings it
>can play. I know you try to guide it into openings you think it plays well, but
>it does not actually get a vote in the decision. This is a qualitative
>difference between what you do and what a second does.

If you are really interested in this question, a great book to read is Jonathan
Schaeffer's - One Jump Ahead, Springer 1997.

To quote just one paragraph (and the only one that is sufficiently on-topic
here):

"A third approach was to have someone filter the checkers literature to identify
what was important and manually enter that information into Chinook. Obviously,
this wasn't desirable because of the extensive commitment of human time
involved. But there was a precedent. To solve the opening problems of his chess
machine, Belle, Ken Thompson typed in opening lines from the "Encyclopedia of
Chess Openings" (in five thick volumes). Religiously, he dedicated one hour a
day for almost three years (!) to the tedious pursuit of entering lines of play
from the books and having his Belle computer verify them. The result was an
opening library of roughly three-hundred thousand moves. The results were
immediate and obvious. Belle became a much stronger chess program, and Ken
probably aged prematurely. Later Ken developed a program to automatically read
the "Encyclopedia", allowing him to do in a few days what had taken him three
years to do manually. " (Henry Baird and Ken Thompson, "Reading Chess" , IEEE
Transactions on Pattern Analysis and Machine Intelligence, 12, no 6(1990) : pp.
552-559

Just the thought of someone like Ken Thompson entering opening lines into his
chessprogram for three years, gives me the shivers ;).

The book also adresses the philosophical and maybe even moral questions in later
chapters.

Peter



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.