Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Garry still singing the same Deep Blue blues...

Author: blass uri

Date: 23:50:08 05/07/00

Go up one level in this thread


On May 08, 2000 at 01:01:36, Pete Galati wrote:

>On May 08, 2000 at 00:29:15, blass uri wrote:
>
>>On May 07, 2000 at 21:47:26, Robert Hyatt wrote:
>>
>>>On May 07, 2000 at 17:14:25, Hans Gerber wrote:
>>>
>>>>On May 07, 2000 at 16:16:07, Ed Schröder wrote:
>>>>
>>>>>There are 2 kinds of science, the one about money and the orginal one
>>>>>which is about complete openness. Guess about which one I was talking :)
>>>>>
>>>>>Ed
>>>>
>>>>I think you both are talking about different things. Let's take a look. It was a
>>>>test for the strength of the machine. It was not about the question if the
>>>>machine _and_ its helpers could outplay Kasparov psychologically.
>>>
>>>
>>>No it wasn't.  The question, clearly asked, was:
>>>
>>>  "Can a computer beat the world champion in a match at tournament time
>>>   controls?"
>>>
>>>Nothing more, nothing less.  Not "Can the machine work its way thru the
>>>candidates matches first?"  Not "can the machine beat the world champion
>>>after playing a few hundred public games?"
>>>
>>>It was _all_ about beating the world champion.  No-holds-barred.  That is all
>>>that was _ever_ asked.  If you want to ask a different question, feel free.
>>>Perhaps someone will try to answer it.  But for Deep Blue, the question that
>>>was being asked was as given above.  Not with all the qualifiers that Kasparov
>>>wanted to add after the fact. (I want program output.  I want more games.  etc.)
>>>
>>>>
>>>>Scientists constructed the machine. As I said they forgot the question of the
>>>>control of the output of the machine.
>>>
>>>What you ask for is _impossible_ to provide.  That is the very nature of
>>>computer software.  It is very easy to change.
>>>
>>>
>>>
>>>
>>>
>>>>
>>>>Scientists have a certain reputation no matter if they act in science, in a test
>>>>or in a match.
>>>>
>>>>By upsetting Kasparov they violated their own standards. It was not just the
>>>>denial to provide Kasparov with the prints of the output, it was the way this
>>>>has been done. Very unusual behavior for decent scientists.
>>>
>>>Hmmm.. did you _see_ the press conference after game 6?  Would that make _you_
>>>want to cooperate with your opponent?  After he called you a cheater?  After
>>>he implied DB had 'help'?
>>>
>>>They didn't violate any standard.  They were trying to answer one specific
>>>question, which they did.  Not trying to answer other questions that people
>>>want to get the answers to.  Just the one question at the top of this post.
>>>
>>>
>>>
>>>
>>>>
>>>>It was said that the IBM officials and not Hsu et al. were responsible for that
>>>>indecent behavior.
>>>
>>>
>>>Who was responsible for Kasparov's indecent behavior?
>>>
>>>
>>>
>>>
>>>>
>>>>Anyway, the scientists and chess experts around Hsu should then have protested
>>>>against such a method. Because the results of the match would no longer be
>>>>valid. They should have explained that IBM had invested so much money for
>>>>nothing if they treated Kasparov in such an unfriendly way.
>>>>
>>>
>>>After the way he acted, IBM responded _exactly_ in the same way I would have
>>>responded.  I would not have said "no".  I would have said "hell, no" to his
>>>request to see the output.
>>>
>>>I don't think IBM was unfriendly at all.  They refused a ridiculous request,
>>>made in an insulting way.
>>>
>>>
>>>
>>>
>>>>I don't understand why these scientists gave their consent to the destruction of
>>>>their own experiment and probably their own reputation.
>>>
>>>
>>>
>>>Their reputation is doing just fine.  DB wasn't destroyed.  The chess processors
>>>are sitting on a shelf.  Newer SP machines are rolling off the assembly line.
>>>But if I were them, I wouldn't play Kasparov again.  He acted like a jackass.
>>>Let 'im stew in his own juice after the antics at the various press conferences,
>>>and the public statements he has made since the event.  He doesn't deserve a
>>>'rubber match' with them.
>>>
>>>I don't even think he really wants one.  He refused an offer from Hsu already.
>>
>>The problem is that Hsu only wants to play with Kasparov.
>>
>>I believe that he can sell many deep blue chips if it can beat one of the >2700
>>players that is not Kasparov in a 6 games tournament time control match.
>>
>>Uri
>
>I would like to see Deep Blue go up against some of the other high rated GMs,
>but not just one, I'd like to see more than one match that is.  On the other
>hand, I don't believe it carries as much weight and selling power to face the
>other >2700 players as it does to play Kasparov.  I think that's unfortunate.
>
>Pete

I think that if Deeper blue can get performance of more than 2800 in a match
against one of the best 10 players in the world then the name of the GM is not
very important for selling the progam.

I think that most of the buyers of chess programs like Fritz and Junior will buy
Deep blue for 200$ if it can beat players like anand 4:2 at tournament time
control games or if it can beat programs like Fritz and Deep Junior 90:10 in a
tournament time control match.

I may be wrong because I do not know what other people think.

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.