The March of the Machine

Ever since I got my copy of Fritz, I have renamed computer chess programmes collectively as The Destroyers of Dreams. That Bxb5+ sacrifice which took you ten minutes to spot, another ten minutes for your half-shaking hand to play, and soon a crunching victory was yours!- you come home grinning, but twenty seconds later the computer flashes up the forced draw your opponent failed to find, you too in the post-mortem after. Or you arrive home thinking, I positionally outplayed him, it went from equal to plus-equal to +/- and then I mastered that endgame; such a nice, smooth game. But the evaluation graph shows mountains and valleys, up and down like a hyperactive yo-yo - for in the midst of your vague shuffling, each of you missed dramatic pawn breaks all over the board. Your smooth game has been turned into the battle of two blind toddlers, each armed with drooping swords, dripping with rust.

Personal vanity aside, chess computers are busy destroying broader chess dreams too. The most conspicuous is the idea humans are good at it. That with our intuition, imagination, fighting spirit, understanding, we transcend brute force and formulas and at our best, make an art out of a game. Of course - we can make excuses for each of mankind's losses to the computers - Kasparov was distracted by the possibility IBM were cheating, Kramnik's mate-in-one blunder was a total freak, and how it all might have been different, and so on. But the fact is, we have to make these excuses every time, not just as a one off.

The latest not-just-a-one-off, which finished last week, was nonetheless an interesting encounter. As previewed, the computer Rybka would take on Grandmaster Jaan Ehlvest with considerable odds. An opening book only going to move 3. No tablebases. Black in each game. Half the starting clock time, a third the increment. Quite good odds for the human, you might think?

Not really, as it turned out. Ehlvest drew three and lost three. You can find the often-interesting games, a brief report and comments here on the Rybka Forum, whilst hopefully Ehlvest will update his blog with more comments than currently there. Finally, there was a particularly interesting interview at chessvibes with Rybka's creator Vasik Rajlich and Larry Kaufman, who organised the match. Here's one choice quote from many:

any match between a human and Rybka where Rybka doesn’t give a material handicap must be played with a tiny Rybka book just to keep things reasonably competitive. Even this doesn’t seem to be enough, so probably, we will look in the direction of material handicaps now.

Perhaps rather than destroy dreams, a different dream is emerging: that the Romantic, amateur, café-based era of chess will be reborn - but instead of a human defeating ten opponents at once, each at pawn-odds, and with a game of whist on the side, a little laptop performs this miracle instead, and the line of patzers-victims consists solely of super-Grandmasters.

Or is that a new chess nightmare?




PS, for our UK readers. If you prefer chess on TV to chess with computers, there is a programme called "Make Me a Genius", part of the My Brilliant Brain series, on Channel 5 tonight from 9pm to 10pm. It's all about Susan Polgar, and The Radio Times says: "This episode focuses on 38-year-old Susan Polgar, the first female chess grandmaster, whose incredible story suggests that genius does not always have to be innate, but can be taught. How has Susan trained her brain to such a formidable degree?" The rather odd thing here is that Susan Polgar is not a genius, so her story actually suggests the exact opposite conclusion: no matter how much training someone has, you need genius to be a genius. Thanks to Angus for the tip-off.