Author: William H Rogers
Date: 14:09:10 04/14/99
Go up one level in this thread
With all of the maga bytes that the newest programs have now, it is a matter of time before chess might be solved. For example: some programs have opening books that go 20 moves deep, and others have ending data bases that can see a mate in 10 or more moves. There have been some games played where one person has been checkmated without the program ever having to leave its books! Sooner or later someone will be able to put it all together and find a solution to most games, but I think that it will only be accomplished when someone has finally figured out a different kind of chess knowledge to program in. ie. something besides tactics, meaning almost explictly captures. When that person finally puts a handle on positional plays, then we will see a great new game. It was once said that for every step forward we take, we sometimes we have to back up a couple of squares and start over. I would still like to see chess tournements where the programs are limited to on 10 or so moves in their opening books, and no end game hash tables. Just as soon as that edge is eliminated, I think we will see some better improvements in chess programs. After all if all of the programs share the same books, then the final knowledge will lie in their chess knowledge, but as long as it doesn't, the improvements will come more slowly. At least that is my oppinion. Bill
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.