Author: Dann Corbit
Date: 15:01:28 06/21/01
Go up one level in this thread
On June 21, 2001 at 17:51:26, Mark Young wrote: [snip] >Are you saying a Grandmaster can tell nothing about a programs strength, even >when he can study the games? Are you saying Kramnik must be a chess programmer >or a chess program expert to understad how strong computers play? > >I don't understand. On the one hand, nobody can argue that Kramnik is a chess super-genius. A weird, fluky phenom with bizarre chess insight. He can (no doubt) look at a game score and see brilliancies and errors at a glance. On the other hand, most GM's are ignorant about how to play against computers. I have sent some GM's email after high-profile computer matches and asked about their strategy. Generally, they just play them like they would any other opponent. Some are *totally* ignorant of what is hard for a computer to fathom. I *suspect* that if Kramnik plays anti-computer chess after 3 months of perparation he will slaughter the computer. But if he plays it like any other opponent, I think it might be a very even match. On the other hand, I could be wrong about computers. Maybe computers have already advanced so far that even the right strategy is not that great of an advantage any more. Certainly, somewhere during the match any human -- even Kramnik -- will make some tiny tactical slip that will cost some fraction of a pawn. Maybe that could be enough for an 8 CPU top level program to grind out a win. In any case, I know that the advice Kasparov got about how to play against Deep Blue was very, very bad advice. In general, the knowlege level among GM's about how to face computers might have risen a little bit. But not much. It will sure be an interesting match -- any way you look at it.
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.