Author: Pete R.
Date: 09:53:36 04/11/00
Go up one level in this thread
On April 10, 2000 at 20:01:57, Jorge Pichard wrote: >In tournament time control it will take a least another 3 Years for a P.C. Chess >Software to be able to beat Kasparov in Tournament time control, only if by then >either Intel or AMD organize an interesting Chess match which will encourage >Kasparov to play against the best P.C. Chess Software then. I strongly believe >that at least it will take a Dual 3 Ghz SMP To beat Kasparov. I am quite sure that Kasparov has been "beaten" many times by his laptop, whether it's running Fritz or Hiarcs or Junior. That is, he tried a move he thought was good, and the program refuted it handily. That's the normal process of working out lines and novelties with the computer, which he does a great deal. We already know that computers are superior at G/25 and similar controls. My hypothesis is that the only way top humans can beat computers at tournament time control now is that they know they are playing a computer, and so they play to the unique weaknesses of the programs and avoid their strengths. If the top players thought they were playing humans, I believe they would get crushed more often then not, and therefore the computers would have the higher ratings. I work through the top players' games all the time with computers, and they make mistakes that a program would punish mercilessly. I can't prove this hypothesis since the required test scenario (fooling human GMs into thinking they are playing other human GMs) would likely piss them off. ;) In any case I believe the computers would prove they are already superior at tournament controls under such conditions, and top players currently beat them only by playing very conservatively and attempting to address their known weaknesses.
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.