Author: Howard Exner
Date: 23:55:56 04/22/98
Go up one level in this thread
On April 22, 1998 at 21:34:39, jonathan Baxter wrote: >Far more interesting than how much better DB is against other computers >is "How much better is DB than humans". DB "thinks" in more or less the >same way as other programs, yet deeper, so its going to be stronger---so >what. But against the top humans, well, I for one was not convinced by >the DB-Kasparov match. The only conclusion I drew from the match was the not so profound one that, "Deep Blue won the 6 game match". Like you I refrained from the conclusion "Deep Blue must be better than Gary". I could not conclude that Deep Blue was yet a Super-Grandmaster. Many more games are needed for that assertion. I was impressed though by the remarkable defense Deep Blue put up in games 3,4 and 5. That must have been disheartening for Gary. >If DB was in open competition, the Super-GM's >would quickly find weaknesses in its play, and probably at a rate that >even IBM's large team could not keep up with fixing. My guess would be >that after a while DB would struggle to maintain a rating over 2650 >because the really strong human players would be really well prepared >gainst it. Yes, familiarity breeds not only contempt but rating points too. I think Comp vs Human games will remain as fun events as in Anand-Rebel and Crafty-Yermolinsky (on the internet). Sure wish the Harvard Cup or something similar would emerge again. As for serious rated play in FIDE tournaments it will never happen. I believe computer play would be to closely defined to resemble a form of correspondence chess for them to be allowed the chance at a GM norm.
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.