Author: Dann Corbit
Date: 18:56:55 03/15/01
Go up one level in this thread
On March 15, 2001 at 21:26:27, Terry McCracken wrote: >On March 15, 2001 at 19:59:08, Sonja Tiede wrote: > >>On March 15, 2001 at 06:31:52, Terry McCracken wrote: >> >>>On March 15, 2001 at 04:47:29, Ulrich Tuerke wrote: >>> >>>>On March 14, 2001 at 19:46:11, Sonja Tiede wrote: >>>> >>>>2800 +/- 600 >>>That absolutly makes _No_ sense whatsoever. It was Never as low >>>2200 and never as High as 3400. >>> >>>Deep Blue was close to 2700 +/- _50_ FIDE and that was about it. No great swings >>>in it's play like you suggest. Deep Blues' play was fairly balanced >>>and was about 2700 FIDE. It did of course have amazing tactical albility >>>based on search wich would _appear_ to make it look 3000+ in some situations >>>and much less in very strategic positions, which closely combined strategy >>>and tactics. Game One was a good example, and Kasparov _crushed_ it. >>>While Game Six was _worthless_ as GK mixed up his move order on move 7 while >>>in book. He for some reason thought he had played 7...Bd6, and on move 8 Black >>>plays h6. >>> >>>T.M. >> >> >>I agree with your opinion. (2700+- 50) >> >>DB has played 12 games against Kasparov (2 matches) and its total score >>was less than kasp. (the improvements from J.Benj. were small, because you >>cannot impove the strength of a program easily by putting >>a GM in front of the machine, telling him to make changes on some parameters) >> >>Sonja > >Thanks Sonja, and I concur with your view as well. IBM took their _fortunate_ >win of the second match and ran! I see the opposite. Kasparov whined like a petulant child. In hindsight, I am sure we can find mistakes by anyone, including both Kasparov and Deep Blue. But under the tournament conditions, he was clearly beaten both OTB and mentally as well. >Truly it was a " Sad Day " for chess, chessplayers, and chesscomputers! It was a banner day for AI. It was the most famous day in the history of mankind for chess -- far more famous than (for instance) Kasparov verses the world. It was the finest day in history for chess computers. >IBM made excuses for packing up the 8 year project, saying they were done >and moving on to other horizons. That's because they were done, and moving on to other horizons. Was anyone really so foolish as to believe they spent millions of dollars for something other than a return on their investment? They have nothing whatsoever to gain from another match and everything to lose. Of course we won't see another like it -- nor would any reasonable person expect to see one. >This should have been an ongoing scientific project but they just pulled the >plug. You are welcome to fund a similar project. If you offered Hsu a few million, he might be interested. >It's 4 years later and still no one has picked up where IBM left off, and I >don't expect anything close to the Deep Blue Project anytime soon. Cost/benefit analysis shows it would be silly to do so. Nevertheless, in 3 years, a 64 CPU Alpha machine using next generation chips will be able to do in software what Deep Blue did in hardware. In ten years, maybe your desktop computer will. In any case, we are talking about multi-million dollar machines here, where CPU time is very expensive and it takes a large team to operate it. >Poor science, but good marketing from IBM. Poor science? They beat the best player in the world in 1997. Nothing else on the planet at the time had a prayer of doing so. >Still that is dabatable as well, >there was money still to be had with the DB but IBM just did'nt have the right >plan after their serendipitous success. They had the most sensible imagineable plan. It is exactly what everyone should have expected. If they expected otherwise, all I can say is: "What in the world could they possibly have been thinking?" They got tremendous publicity/PR/exposure/etc. They are still selling that image now four years later -- and quite successfully. >Terry McCracken IMO Deep Blue is one of the modern wonders of the world. IBM did nothing wrong. To imagine that they have some sort of duty to entertain us is a fairy-land vision completely disconnected from any sort of reality.
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.