Looking Back at Deep Blue
Part of Deep Blue on display. (Source: Smithsonian)
For all the talk of the current boom in artificial intelligence, it is easy to forget that these advances are a long time coming. Perhaps the most memorable watershed moment in the history of AI was the 1997 chess match between Grandmaster Garry Kasparov and Deep Blue, an IBM-built supercomputer. Deep Blue beat Kasparov, considered the best chess player in the world, in the final game of a six-game rematch, a major legitimization of artificial intelligence.
Kasparov and Deep Blue first played in 1996 in Philadelphia, where the machine took one game, Kasparov took three, and the other two were played to draws. A year later, after Deep Blue had scrimmaged Grandmaster Joel Benjamin and others, they played again in New York’s Equitable Center skyscraper. Kasparov won one game, Deep Blue another, leaving three draws, with the final game determining the winner. Kasparov ultimately resigned the game, handing the victory to Deep Blue, though he later accused the computer of being aided by a human Grandmaster, much like the 18th century Mechanical Turk machine.
Kasparov playing Deep Blue during the 1996 match. (Source: AP)
Kasparov’s charges of fraud were never proven—Deep Blue was simply the product of many chess computers that came before it. It was a successor to IBM’s previous Deep Thought (named for the computer in Hitchhiker’s Guide) and laid the groundwork for Deep Junior, Deep Fritz, and others. The machine operated 480 processors across two computer towers, capable of calculating 200,000,000 moves in a single second. With this approach, the computer does not have to think so much as it has to simply run the numbers and brute force its way to a victory.
Today, chess computers can trounce even the most adept Grandmasters and the entire field is something of a settled science. Deep Blue’s victory is hailed as a major step forward for artificial intelligence, demonstrating the capacity for AI to function in a fixed system like a chess board. However, perhaps it should be remembered that Kasparov was able to hold his own for so long because he did not use brute force, but drew from abstract reasoning and experience to make his moves, something a computer cannot reproduce in the same way.
Newborn M. Beyond Deep Blue : Chess in the Stratosphere. Springer; 2011. doi:10.1007/978-0-85729-341-1