Author: martin fierz
Date: 14:40:33 10/24/02
Go up one level in this thread
On October 24, 2002 at 08:05:41, Jouni Uski wrote: >I mean which result would be DEFINITE/FINAL/LAST etc. indicator, that computer >is better than best human in chess. >10-0? 20-0? 40-0? And may be human mated in under 30 moves in each game? Should >be quite clear then for most, if not for all. > >Jouni mated in under 30 in each game will never happen - you just can't do that. and the result does not have to be 10-0. personally, i want to see that the machine is capable of outplaying the human. and not just scoring on blunders. this is something which rarely happens, e.g. in the DF-kramnik match it *never* happened. it has happened though, for example DB-kasparov, game two of the rematch; rebel-van wely, game 3; hiarcs-gulko; smirin-hiarcs come to my mind. i do not doubt that in the future we will see this happening more often. as it is now, computers mostly just "play along". it's like winning a game of tennis against "unforced errors" of the opponent. that is different than winning it by plaing lots of great winners. if we see a match which the computer wins against a top GM where it actually outplays the human, then i will be convinced. the score can be only 6-4 or something low, but the games are the "proof". this concept is obviously rather subjective, unfortunately... aloha martin
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.