Author: Randall Shane
Date: 13:35:08 10/16/02
Go up one level in this thread
On October 16, 2002 at 15:34:40, Uri Blass wrote: >On October 16, 2002 at 11:52:09, Randall Shane wrote: > >>On October 16, 2002 at 11:19:47, Uri Blass wrote: >> >>> >>>There are 3 possibilities: >>> >>>1)Hsu is not interested in making money >>>He is interested only in beating >>>kasparov or kramnik. >>> >>>If deep blue chip can beat all >>>the top programs of today then I expect >>>people to buy it(but people need first to >>>see a proof by a match against the top programs >>>of today when Hsu does not have the opponent). >>> >>>2)Hsu does not believe that he can make money >>>from a new chip or IBM does not allow him >>>to make money from a new chip(I do not >>>understand a lot of details >>>about it for example why does IBM forbid him >>>to discover all the details about deep blue >>>evaluation) >>> >>>It make IBM looks bad and it seems that they >>>have things to hide. >>> >>>3)Hsu knows that he cannot create a chip that >>>is better than the top programs of today. >>> >>>Uri >> >>Disclaimer -- I work for IBM. However, I don't do chess for IBM, I don't >>anybody who worked on that project (or even anybody who knows anybody!), and the >>closest I've been to Watson Research Center is driving past it on the way to a >>Japanese steakhouse. I'm probably less qualified to tell you what IBM is >>thinking than your average Magic 8-Ball. Nevertheless... >> >>I'd think that it's a combination of #2 -- IBM didn't get to where it was by >>releasing information of any kind without a good reason -- > >If IBM does not plan to continue to play with Deep blue then I see no reason for >them to force Hsu to keep details secret. IBM is a large corporation that has its own reasons. IBM's purpose (not Dr. Hsu's purpose!) after the second Kasparov match was to maximize publicity and potential future revenue. Releasing internal evaluation details doesn't help that. Doesn't hurt it either, I would think, but since it doesn't help -- it's not been released. An air of mystery has probably served IBM better than tons of details would have, I suspect. Also, maybe IBM wants to sell that data someday. When trying to figure out why a large corporation does something, whether the corporation would look bad or look good, and whether the corporation would look open or look like it's trying to hide something is usually irrelevant. Look at the bottom line, the dollars and cents (or at the euros, or the yen, or the pounds sterling...). >If they keep details as a secret then it is natural that people suspect that >their evaluation was not good not only because of lack of tuning but also >because of lack of knowledge. Hmm, Dr. Hsu has released far more details about Deep Blue's internals than Fritz's developers have released about Fritz's internals. I certainly can't understand why Dr. Hsu gets the crap while the Fritz developers get the praise. > and a #4 you're >>forgetting -- Dr. Hsu may just be really tired of it all. With all the >>unjustified crap he's taken in and out of the computer chess and business >>communities, and the fact that he's spent nearly all of his adult life doing >>computer chess, I can imagine where he may just want to Do Something Else. He's >>written a book, documented his life, maybe he just doesn't want to do it >>anymore. > >The crap is result of the decision to leave computer chess and at the same time >say that deeper blue is better than Deep Fritz. The crap flung Dr. Hsu's way started seconds after Deep Blue II beat Kasparov (well before he decided to leave the field) and has apparently continued unabated to this day. This is not to say Deep Blue was perfect -- it wasn't. But it was good enough to beat the World Champion. There's no particular reason to believe that it's evaluation sucked, although it's certain that it could have been improved. > >If he does not want to continue then he should say nothing about deeper blue. > >Uri Deep Fritz's team can talk, but Dr. Hsu's gotta shut up, eh? Ed Schroeder is supposedly getting out of active development of Rebel (correct me if I'm wrong, please) -- does this mean that he can't comment on computer chess anymore? Who gets to speak, and who doesn't?
This page took 0.01 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.