Author: Uri Blass
Date: 08:07:39 12/29/01
Go up one level in this thread
On December 29, 2001 at 07:39:37, Tom Kerrigan wrote: <snipped> >I would not trust my program to produce moves that are any better than GM moves. >I'm not sure I'd trust anybody else's program, either. It is not hard to find >examples of programs giving you bad moves after "long" searches. 1)I agree that there are examples when programs give bad moves even after long search but I believe that it is only minority of the mistakes of computers and there are not less positions when GM's give moves that are weaker than computer strength of the top programs and not only tactical mistakes. Computers already did performance of super GM's in some tournaments. 2)Another problem is to know if the computer does a mistake because it is possible that humans may believe that a move is a mistake when computers after long search know better and it is also possible that there is more than one good move and humans who find a winning move do not care to check that it is the only winning move. Humans may try to check every possible move by analysis with computer but it takes time. Uri
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.