Author: Mike S.
Date: 12:20:17 04/12/02
Go up one level in this thread
On April 12, 2002 at 14:11:40, Roy Eassa wrote:
>(...)
>PPS: All it will take is ONE non-GM human player who can consistenly beat a top
>program to prove that top programs are inconsistent (at best) in their ability
>to play at top-GM level. AFAIK, has there NEVER been a non-GM human player who
>could consistently beat a top GM.
I think Eduard's games are great achievements. I know how difficult it is to go
for the point successfully even after the program has captured the poisoned
material. Eduard posted a training position once, and it was extremely
difficult, actally impossible, for me to find a winning continuation against
Fritz (I tried for hours). But there was one.
The point is: It *is* possible to exploit that type of weakness, but that's much
too risky for a match under standard conditions. The method is very
opening-dependant. I don't think you can prepare something like that with White
*and* Black, and for every possible major opening. It's a tightrope walk also.
That's why GMs rely on other types of anticomputer strategies, if at all.
Btw. the quotes from the Krabbé site ("no more significance than a match between
a cat and a book") show a typical misunderstanding of computerchess IMO. I think
what makes computerchess interesting isn't the artificial charakter that much,
but the fact that can do something like a human being can, mainly. In other
words, the fascination comes from the "human side" of the machine so to speak.
For example, we can see computer moves which have the *beauty* of Capablanca's
endgame techniqué. I have put an example below. Krabbé doesn't see that. But the
ability to play chess, even strong chess, is basically a human one at first, and
we are still fascinated (and probably ever will be) that a "machine" can do
that.
Regards,
M.Scheidl
[D]5k2/6p1/2p2p2/P7/1Q6/2P1pqPP/7K/8 b - - 0 67
White has just played Qb4+.
1...c5!! A wonderful example of master-like endgame techniqué (played by
Crafty). The sacrifice has a reason, or idea behind it which looks absolutely
human-like IMO. Furthermore, think how strong a human player must be, to find
that idea here instead of a simple King move out of check. The pawn seems to be
sacrificed out of the blue, whith no appearant reason at a first glance.
Great.
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.