Brain vs silicon battle resumes

There will be lots of emotion on both sides of the board: Kasparov’s unrelenting desire to win and the aspirations of the computer scientists who have devoted years of research to creating the hugely powerful calculating machine.

“Right now we are in an age where we are trying to understand our relationship with computers,” Joseph Hoane, the Deep Blue team’s codewriter, told Reuters. “The computer has shown me things that I could never have known otherwise. It’s a tool that helps us think better, that’s one of the main reasons why we’re here.”

Kasparov, only 34 years old and world champion since the age of 22, is considered by most experts to be the best player of the ancient game who has ever lived. He faces a challenger that has twice the calculating speed of last year’s programme. Of even greater significance, Deep Blue’s software has been infused with more knowledge about chess.

IBM hired US Grandmaster Joel Benjamin to work full-time with the team of five scientists in preparation for the match. Deep Blue can examine hundreds of millions of positions per second and it now also appreciates strategy and the relative value of pieces in certain positions, according to IBM.

Deep Blue is an IBM RS/6000 SP parallel processor with specialised microchips for chess. Parallel processors make many complex calculations simultaneously, unlike serial processors.

Computers have beaten Kasparov and other leading Grandmasters in matches with speeded-up games of an hour, 30 minutes or less, but not in classical chess, where games can last as long as six hours. Kasparov’s game one defeat in Philadelphia was historic — the first time a world champion had lost to a computer programme in the traditional format.

The burly Russian Grandmaster, who has no advance information of Deep Blue’s ability, has labelled it an “alien opponent”. He says he will accept any combination of wins, draws or losses that produces a winning result and enable him to collect the $700,000 prize. The loser receives $400,000.

“Any win. One win, five draws. That’s good, it will do,” said Kasparov, who feels the match is important not only to himself and to chess but also to the wider world.

“Mankind is at a stage now where we need some practical results from our computer work and you can only get this specific result in chess,” he said.

Chess is considered ideal for computers because the game involves a specific number of physical objects governed by simple, clearly defined rules. All of those involved in the match have described it as an experiment to help build computers that can make complex, simultaneous calculations at high speeds.