Deep Blue Vs. Kasparov
mrf...@yahoo.com
* Chess champion Garry Kasparov's latest interactive stunt seemed like a
good idea -- until Deep Blue bested him yet again. Rick Chandler reports
http://www.impressionmag.com/dreams.shtml
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
Lester V. Tinnin
>A funny take on Kasparov vs. the world
>* Chess champion Garry Kasparov's latest interactive stunt seemed like a
>good idea -- until Deep Blue bested him yet again. Rick Chandler reports
>http://www.impressionmag.com/dreams.shtml
>
Kasparov is making money, isn't he. That doesn't sound too dumb. I
am wondering about the computer. Kasparov won playing white. If
Kasparov starts off with the same first move when he has white again,
is the computer going to make the same move as before? The computer
is going to have to make a different move than before sometime into
the game if Kasparov makes all of his former moves because the
computer lost before. But if Kasparov makes all the same moves as
before when is the computer going to blink and change moves, and is
someone looking over its shoulder telling it to blink?
Just a curious thought. I guess we will find out.
Lester Tinnin
Greg Kennedy
>>* Chess champion Garry Kasparov's latest interactive stunt seemed like a
>>good idea -- until Deep Blue bested him yet again. Rick Chandler reports
>>http://www.impressionmag.com/dreams.shtml
>>
>Kasparov is making money, isn't he. That doesn't sound too dumb. I
>am wondering about the computer. Kasparov won playing white. If
>Kasparov starts off with the same first move when he has white again,
>is the computer going to make the same move as before? The computer
>is going to have to make a different move than before sometime into
>the game if Kasparov makes all of his former moves because the
>computer lost before. But if Kasparov makes all the same moves as
>before when is the computer going to blink and change moves, and is
>someone looking over its shoulder telling it to blink?
The DeepBlue team has retained the services of a brilliant
openings expert to remedy this problem for future games.
I can't reveal everything, but in the case of Garry's clever
trap: 1.e3! d5 2.c4!, DB is now prepared with a fantastic
defensive improvement to replace the hideous ...dc. Yes,
after countless hours of intensive research, I was able to
determine that 2. ...e6!! holds the fort, allowing the machine
to actually reach a playable middlegame. Don't ask why
Joel Benjamin (a former GM) was unable to see this, or why
IBM had to fire him and have FIDE strip him of his title. The
main thing is, DeepBlue is now ready to clobber Kasparov
again, thanks to my brilliant openings research and a tenfold
hardware speedup. Our best estimate of the new improved
DeepBlue's playing strength is 2900+. This may seem
outrageous at first glance, but keep in mind the severity of
the handicap of letting Joel determine the opening moves
before "turning on the power" of the machine. We are now,
er, I mean DeepBlue is now, virtually invincible.
- Greg Kennedy
Ben Leostein
ahead of weaker Grandmasters comparing Deep Blue to a regular PC chess program.
If the toughest program out there is HIRACS which plays at about 2600 FIDE on a
200Mhz, 64M of RAM with considerable hashtables, Deep Blue was only estimated
at 2705. Since Kasparov was able to beat and draw the silicon beast on some
matches, and Benjamin and Seirawan have been known to lose to top PC programs
as well (not many times, but a few), can this compare?
Is Kasparov that much further ahead of other players such as Joel Benjamin and
Yasser Seirawan that they couldn't compare either?
Would Kasparov against Seirawan be fair to compare it to Deep Blue VS. HIRACS?
Regards,
Ben
Lester V. Tinnin
Does this mean that Deep has seconds? A human decides the first move?
Or after Deep Blue loses one that these seconds move in and change his
insides? It should be a no holds barred tournament, in case it isn't.
The computer and Kasparov in the "ring" until all the games are
played. Then if Deep Blue needs to be worked on, O.K. But no working
on him during the tournament. Maybe I am off somewhere, but it sounds
good to me.
Lester Tinnin
PMG
Ben Leostein wrote:
If this is still a discussion about why didn't Deep Blue compete in WCCC let me
remind everyone that that is never going to happen. If IBM decided to enter that
arena they would not do it with the exact same program, it would have been revised
many times by now, and the computer that they ran it on would be very different. We
are talking about the reputation of a very major reputable corporation.
As far as all these powerful programs, their game play is full of holes, their King
safety can be shredded easily. None of them are really up to the standards of play
of a strong GM. This doesn't mean that _I_ can beat them, far from it, my favorite
program at the moment is Ychess programed in assembler language by someone named
Roman Antonczyk in Austria and I never come close to winning a game with it..
As far as GMs loosing to top PC programs goes, of course it happens, but it doesn't
happen all of the time. GMs are much stronger than we give them credit for.
Pete
Ben Leostein
>doesn't
>happen all of the time. GMs are much stronger than we give them credit for.
I know GMs are very strong, as I know quite a few of them personally. :-)
Regards,
Ben
PMG
Ben Leostein wrote:
I don't have that privilage, I'd say that you are lucky.
Pete