Skip to content ↓

Here & There

Many resarchers and students at MIT closely watched the historical chess match between Garry Kasparov and Deep Blue, IBM's supercomputer, and a number of them were asked for their observations by journalists.

Sherry Turkle, professor in the Program in Science, Technology and Society, remarked in the Arizona Republic (May 13) on a change between this year's match and the 1996 contest in how most people--perhaps including Mr. Kasparov himself--saw Deep Blue. "The difference this time is there is a greater public recognition that the computer played more like a person," she said.

Public perception of IBM has become somewhat more sympathetic since Kasparov's defeat, she added. The perception of Deep Blue's maker as "more organized, more neural, not a clunky kind of computer intelligence. really has been facilitated by this match."

Charles Leiserson, professor of computer science and engineering, said in the Boston Herald (May 13) that the match results did not presage a takeover of the world of chess by marching automatons. Deep Blue's victory was most relevant "if you viewed it not as a competition with the machine, but in cooperation with the machine," he said.

Professor Leiserson, whose own chess-playing computer, Cilkchess, can analyze two million possible moves per second, saw Deep Blue and IBM's triumph as a challenge and an opportunity. "If I had sufficient hardware to run, we'd give them a run for their money," he said.

Marvin L. Minsky, Toshiba Professor of Media Arts and Sciences, saw Deep Blue's victory as a good demonstration of progress in computer intelligence but hardly a threat to even a small child's questions.

"Nobody knows how to get a machine to understand even a simple children's story," he told The Boston Globe (May 13). It is still far beyond the capacity of any computer, he said, to be able to "respond intelligently" to a plot-based question such as, `Why did Jane do that?'"

Rather than a giant leap for computer-kind, Deep Blue's victory represents a limited step in a slow, steady progression from Alan Turing's first chess-playing program of 50 years ago, Professor Minksy said, adding that he would "not be surprised if, 100 years from now, an old IBM 360 (a 1960s computer) could beat any human opponent" if it contained the right software.

As for Kasparov's suspicions that Deep Blue had been rigged, William Neveitt, a graduate student in electrical engineering and computer science studying artificial intelligence, commented to the Gannett News Service, "this isn't like crawling in with a screwdriver and being out of there in five minutes."


A version of this article appeared in MIT Tech Talk on May 21, 1997.

Related Topics

More MIT News