Author: Uri Blass
Date: 09:02:11 04/02/01
Go up one level in this thread
On April 02, 2001 at 11:32:20, Vincent Diepeveen wrote: >On April 02, 2001 at 10:35:29, Robert Hyatt wrote: > >>On April 02, 2001 at 00:28:42, Uri Blass wrote: >> >>>On April 01, 2001 at 18:42:27, Vincent Diepeveen wrote: >>> >>>>On April 01, 2001 at 11:00:11, Robert Hyatt wrote: >>>> >>>>>On March 31, 2001 at 21:13:22, Torstein Hall wrote: >>>>> >>>>>>I do not think IBM stand to loose anytihing. The big public has forgotten that >>>>>>Deep Blue ever existed! Anyway, I feel the marketing value of DB is low nopw, so >>>>>>perhaps someone with a few extra bucks to spare can buy DB now? >>>>>> >>>>>>Torstein >>>>>> >>>>>>Was it really a monster playing chess named DB? >>>>> >>>>> >>>>>I think you are _way_ wrong. I just got back from a visit to my home >>>>>town with a population of about 1,500 people. I wore a T-shirt that Compaq >>>>>sent me (they took Crafty + an alpha, to the Linux expo, and let anyone play >>>>>it). The shirt had "I survive the compaq computer chess challenge" on the >>>>>back. Several people asked me "hey, how would your program do against IBM's >>>>>monster?" >>>>> >>>>>People remember deep blue. Because _I_ didn't mention it at all. If you ask >>>>>100 random people about the best chess player, human or machine, more people >>>>>will remember the name "deep blue" than "kasparov". >>>> >>>>Exactly. >>>> >>>>So if they lose a match they might lose a market share and a lot of >>>>'deep blue' quotes. The possible market share they might lose is >>>>worth 37 billion at wall street and it's worth a lot of PR too. >>> >>>I do not believe it. >>>Bruce explained that your math is wrong. >>> >>>Uri >> >> >>I don't believe IBM would _lose_ a thing. They would definitely gain more if >>DB won a third match. They would not _lose_ if it didn't, they just would not >>gain as much useful publicity. > >You hit the right word here Bob, > >IF they win such a match then they definitely will not lose a thing. >Definitely true. > >Now we can discuss to great length the chance that they win, but there >is obviously a big risk they lose. Personally i would think that >also a match against the worst ambassador for chess in the world, Kasparov >(as he played around 2300 level in that match in 1997 and we all know >the result for the game of chess: "chess is solved say most AI papers"), >will be a clear and chanceless victory for Kasparov. > >And definitely no game will get a 18 move win over Kasparov again, he'll >not be that stupid again. > >Against Kramnik a 18 move lose would be imaginable (as bad chessplayers >you play on a few moves with computer then till you get -5.0 or something), >but a 18 move win is impossible forever. > >Kramnik can win without any preparement, whereas Kasparov needs to play >his own game instead of playing something where he is an absolute beginner, >even compared to my poor amateur rating (2280 FIDE now though climbing >to IM level slowly). 1)I think that weak players are in no position to say that they could play better than kasparov. I suspect that part of the moves that are considered to be positional mistakes are no mistakes because kasparov saw more than the humans who analyzed the games. I will trust more the top programs and when I finish my correspondence tournament I may give the top programs of today to analyze the positions of kasparov-deeper blue for many hours in order to find mistakes in the games. 2)I do not think that kasparov had special problems against computers before 1997. Kasparov beated Deep thought easily when karpov had to work hard in order to win against deep thought and he did it only thanks to a mistake of deep thought in a drawn rook endgame. I also remember that kasparov won together with Fritz3(p90) in a blitz tournament and later won Fritz3 in a match by lines like 1.e3 Kasparov lost against Genius3(p90) 1.5:.5 at 25 minutes per game but later won the same program on faster hardware 1.5:.5 at the same time control. I believe that Kasparov may have better chances today mainly because of the fact that he learned to play better thanks to training with computers. 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.