When Gary Kasparov beat IBM's chess computer in 1989 he arrogantly told the programmers to "teach it to resign earlier". Yesterday, though, the world champion found himself humbled by a 1.4-ton heap of silicone in a victory for IBM's Deep Blue that marks a milestone in the progress of artificial intelligence. It is a depressing day for humankind in general.
But why did the computer win? IBM's Deep Blue first caused a sensation in Philadelphia last year when it crushed the world champion in the first game of their match. "I was rightfully massacred," said Kasparov. Yet he won the series 4-2, striking a blow for the human race at the same time. Game by game, he learned more about the program's style and weaknesses. He levelled the score in the second game, then made two draws and by the final game he was in control.
But Deep Blue's programmers struck back. They revamped the supercomputer, doubling its greatest strength: the ability to search through millions of possibilities for the strongest move. Last year it could scan 100 million positions per second. The new machine which crushed Kasparov yesterday can scan 200 million a second, sometimes 300 million, and can analyse 74 moves ahead – compared with chess masters who typically think 10 moves ahead. Last night's 19-move loss to the computer was the worst defeat of Kasparov's career. Kasparov has never lost a match to a human opponent and this was the first game he had lost in under 20 moves.
Kasparov has said he alone would halt the advance of computers by creativity and intuition. He had several good positions in the match but his confidence was shattered first by the second game, where he resigned in a drawn position, and by the fifth, which Deep Blue drew by remarkable defence.
The new model, Deeper Blue, was fed with more chess software to beef up its strategic and tactical grasp. But the theory was that Kasparov, aged 34, would be able to prove the human mind, with its flexibility and creativity, could still beat sheer calculating power spiced up with chess knowledge. Computers have a huge in-built advantage as they can evaluate so many moves so quickly, and when asked how many moves ahead he thought, the pioneering 1920s chess theorist Richard Reti replied he was usually one move behind. But whereas computers have to plough through every variation in turn, humans can instantly recognise 99.9 per cent of them are nonsense. That is why most experts thought Kasparov would win.