Success can be lonely with no mate to check

DESPITE its dramatic victory over world chess champion, Garry Kasparov, one can't help feeling sorry for the complex tangle of…

DESPITE its dramatic victory over world chess champion, Garry Kasparov, one can't help feeling sorry for the complex tangle of silicon, circuit boards and processors that make up the supercomputer, Deep Blue.

It's all very well possessing the power to calculate 200 million potential chess moves per second. But what is the point when one is incapable of spending even one cent of the $700,000 prize money that results?

That redoubtable Russian, Kasparov, may have crumbled under its superior skill and yet it lacked the wherewithal to goad him with witty repartee such as: "Cheque please, mate".

Worst of all, when the final pieces were played at the Equitable Centre in New York last Sunday, the inimitable Deep Blue had no Deep Pink with whom to celebrate.

READ MORE

Yes, life can be tough when you're a 6ft 5in tall, 1.4 ton, high-performance IBM computer. At least Arthur C. Clarke's HAL 9000, the "more than my job's worth" mechanical star of Space Odyssey 2001, was programmed with bags of personality.

Deep Blue's immediate predecessor was born nine years ago. Deep Thought promptly shocked the chess world by becoming the first computer to defeat a grandmaster in a tournament.

A microchip off the old block, Deep Blue came to "life" a year later at the IBM research centre in Yorktown Heights, New York. In February, 1996, it was pitted against Kasparov in Philadelphia.

The result Mankind 1, Technology 0 - made last week's rematch inevitable.

The IBM team behind Deep Blue began to tweak and fine-tune their charge, doubling its speed and expanding its chess knowledge. The two-tower structure (an IBM RS/6000 SP to be precise) is stuffed with 32 general purpose computers each attached to eight special purpose processors.

In what has been seen as a pivotal improvement, the data from the entire catalogue of grandmaster chess games played over a century was infused into Deep Blue's memory.

No wonder Garry Kasparov - playing not just the brute force of a computer but the ghosts of grandmasters past - appeared visibly frustrated throughout the six-game tournament.

It is a well-known fact that computers, unlike Russian grand masters, are not affected by stress. Neither do they get irked by the trademark black looks and smirks of their opponents. Crucially, perhaps, they do not need to go to the toilet.

But the odds were not all stacked against the Russian; Deep Blue's shortcomings are also well documented.

The computer's lack of intuition, creativity and imagination - often the chess player's greatest weapons is what pundits believe stopped Deep Blue from earning a more comprehensive victory (3.5 games to 2.5).

Defeat was hard for Kasparov to swallow. In a less than noble outburst after game two he more or less accused Deep Blue of cheating.

Chess experts had been surprised when Deep Blue's play took on almost human proportions as he flummoxed Kasparov with a sequence of moves which began with an unexpected pawn exchange on move 36.

The suggestion that Deep Blue may have had a little help from his scientist friends is one some have not totally discounted.

"What Kasparov was saying was that at certain points the computer was playing in a surprisingly human way," says Alexander Bahurin, a Russian grandmaster resident in Ireland. "Certainly, I believe that it is possible that some support was given to the computer during the tournament, especially in areas like long-term strategic planning where humans are known to be better."

Kasparov's defeat is seen as significant in chess circles. It has also sparked a debate around the world.

Some more radical commentators say it is a defeat for mankind, despite Deep Blue having been created by humans. This doom-laden reaction is played down by the scientists in IBM's research centre, where they go so far as saying that compared with humans, Deep Blue is relatively "stupid".

"Deep Blue is stunningly effective at solving chess problems but it is less intelligent than the stupidest person," explains Murray Campbell, one of eight IBM research scientists who can boast a "parental" role in Deep Blue.

So what's next for Deep Blue? Kasparov wants a rematch, while Club Xiang Qi, the Chinese chess organisation, has said that if Deep Blue can beat a human player in Chinese chess it will hand over $1 million to IBM. Kasparov's longtime adversary, Anatoly Karpov, has also issued a challenge.

But the biggest threat for the next generation of Deep Blue may come from a so-far-disregarded sector of the chess playing community. Women's world champion, Susan Polger, has offered to massage mankind's bruised ego in the wake of Kasparov's defeat.

Her secret weapon in a battle with Deep Blue? "Women's intuition," she says. IT has been warned.