Grandmaster Veselin Topalov on chess computers
news24.com had an interview last week with GM Topalov's view on computers and chess:
"At the present moment I don't fear humans. Computers I think are very strong," the Bulgarian grandmaster said.
"The problem against computers is you cannot scare them. This is a big problem, you cannot bluff, (there's) no psychological fight ... The good thing is I don't care if the computer calculates 100 million moves per second, or 200 million, for me it's about the same, I don't see any difference."
Topalov says that in ten years, computers will be better than 99.9% of humans. I think that is the case now (maybe even 100%), and in ten years the game may be completely solved, like checkers practically is today. As you can see from Kurzweil's chart, he predicts that by 2013 we will have the computing power to produce a simulation of a functional human brain.
Hi,
Topolov is arguably the strongest human player curently, but I'm not so sure his opinion is much better than yours or mine.
It is a huge problem if attacked by just brute force calculation, and the singularity notwithstanding, probably insoluble in the time frame where chess is even relevant. Theory is interesting but also having a tough time of it.
The number of chess games has only been shown to be finite if you assume that a player will always claim a draw under the 50 move rule:
Each of the 16 pawns has at most 6 moves before being promoted. there are now 16 * 6 = 96 possible P moves. 30 pieces can be taken during a game. So 30 + 96 = 126. If these moves happen as infrequently as is possible, on the 50th move, the maximum number of moves is 126 * 50 + 50 = 6350 moves for a game of Chess. This is probably lower, but this is an upper bound.
White's advantage is so small that it can't even be shown that a strategy exists whereby white can't lose. In 1913, Zermelo posited that either white can force a win, or black can force a win, or both sides can force at least a draw. That has been proved, but basically that's where theory stops so far as I know.
My suspicion and hope is that with best play on both sides, chess is a draw.
Been reading UC for about a year. Almost took the plunge on CONN. Wish I did. My recent best are NTG and CFK. Worst was PCYC. I know money is money but somehow I'd rather trade and own a company trying make money by curing metastatic lung cancer, than one, say, trying to make it by broadcasting Howard Stern from satalites. So I know I too have it in me to irrationally do real damage to my portfolio.
I will be at the USATE in Parsippany in Feb.
Regards,
JB