| Moment Home Page | Current Article Index | Search | Back Issues |

Deep Blue's Intelligence?

by Brian E. Shicoff

A series of chess matches only two months ago has again brought new life to the question, "Can machines think?"

From February 10th to the 17th in Pennsylvania, Pa., Garry Kasparov, World Chess Champion, won three matches, drew two, and lost one, to the strongest chess playing computer ever developed. Although Kasparov claimed the 400,000 dollar prize for winning the match (the "computer" claiming 100,000), the astonishing fact was that the World Champion lost the first game.

Before the first match even started, Tony Marsland, president of the International Computer Chess Association, is quoted as saying,"Computer chess experts have predicted that within five years the world chess champion will lose a match to a computer...this match will be the first such attempt." After the first game was over, Mr. Marsland, spectators, chess experts, and even Kasparov himself, were shocked that DEEP BLUE had actually played an amazing game and beat the world champion.

In a recent Time article, Kasparov discussed how in one certain position of that first match, the computer offered a pawn sacrifice (usually a sacrifice of material is offered to gain a lead in position). Although Kasparov probably would have made the move himself, he realized that the outcome of it was uncertain. Kasparov stated, "I was stunned by this pawn sacrifice. I had played a lot of computers but had never experienced anything like this. I could feel-I could smell-a new kind of intelligence across the table." DEEP BLUE later went on to recover the sacrificed pawn and, in effect, to win the first match.

In the final five games in which Kasparov drew two and won three, he played with a different style. "...my overall thrust was to avoid giving the computer any concrete goal to calculate toward...[Because it cant fulfill this priority] the computer drifts planlessly and gets into trouble." Kasparov concludes, "So although I think I did see some signs of intelligence, it's a weird kind, an inefficient, inflexible kind that makes me think I have a few years left."

GM Kasparov and Feng-Hsiung Hsu, Ph.d. Hsu was on the team that created Deep Thought, the predecessor of DEEP BLUE
Originally, when Kasparov became World Champion back in 1985 at the age of 22, programmers and computer engineers had been hard at work in constructing solid chess playing machines. But constructing chess playing computers did not start up recently. In fact, the first chess program was written by Alex Bernstein of the Massachusetts Institute of Technology between 1958 and 1959. In 197, the Machack IV computer was the first to play in a human chess tournament. With the introduction of integrated circuits, the first chess playing computers went on the market in 1976. But not until 1983 did computer technology triumph over a chess master in any tournament. A few years later the Deep Thought project was launched, but lost miserably to Kasparov in 1986. DEEP BLUE succeeded Deep Thought and was developed to make up for inadequacies in programming and hardware. Thus, 50 years after the invention of the ENIAC, in the location this first computer was invented, Philadelphia, DEEP BLUE would truly test Kasparov's strength.

The way DEEP BLUE works is as follows: it has 128 processor chips running in parallel with an IBM POWERparallel Systems SP2 as the host computer. Basically, this results in a 1000 times increase in processing speed over Deep Thought, yielding a calculation of one billion positions per second. In approximately three minutes, BLUE could calculate 50 billion positions. Strangely enough though, 97 percent of this computer which an IBM team built was constructed from components that can be purchased by the average consumer.

Despite the strength and numbers of such calculations, however, DEEP BLUE was no match for Kasparov even though it did win the first game. Kasparov, in the three wins he scored against it, showed that the computer could not effectively create plans based on its programming. So maybe BLUE did not have any true intelligence. Maybe it just calculated all the possibilities it could in the time allotted. But maybe, in the not so distant future, a truly intelligent chess computer will sit atop the throne once held by Garry Kasparov.

As for now, however, computers still have a long way to go in becoming truly intelligent.

Internet Chess Club
I.B.M. Home Page

| Moment Home Page | Current Article Index | Search | Back Issues |