Printer Friendly

Chess champion sinks Deep Blue's figuring.

When world chess champion Garry Kasparov faced the chess computer Deep Blue in a six-game match, the contest was not so much man against machine as man against men with machine.

By the end of the match, Kasparov had outmaneuvered Deep Blue's developers-Chung-Jen Tan and his team at the IBM Thomas J. Watson Research Center in Yorktown Heights, N.Y.-to win. But it was a tough battle.

"I did not expect that it would be that tough," Kasparov said after the final game of the match, held last week in Philadelphia. In 1989, he defeated Deep Blue's predecessor, Deep Thought, with relative ease (SN: 10/28/89, p. 276).

After a stunning loss in his first game against Deep Blue, Kasparov adjusted his playing style to exploit weaknesses in the computer's play, winning the second, fifth, and sixth games and earning ties in the other two.

"I think the main distinction between us and computers [is that we] can learn," Kasparov noted. "I learned a lot from game 1 and game 2."

In later games, he deliberately created crowded situations that gave the computer few options, limiting its ability to attack his major pieces.

In contrast, the IBM team could not adjust Deep Blue's program during the match. Late in the fifth game, for example, the computer's calculations indicated that it had a losing position, and it moved pieces in ways that actually made the situation worse.

A human player might have reacted by gambling on a strategy that could trick an opponent into making a mistake. Instead, Deep Blue always assumed that its opponent would play perfectly. The computer lacks a sense of "creative desperation," comments IBM team member A. Joseph Hoane Jr.

A number of glitches probably contributed to Deep Blue's defeat. For instance, the team routinely selects the computer's opening moves, but in game 2, the computer failed to play the chosen opening because someone had stored the information in the wrong place.

Deep Blue's main strength lies in its ability to select the best possible move after evaluating the consequences of various moves far more deeply into a game than a human player can. The games demonstrated that this brute force can go remarkably far in matching human intuition, experience, and knowledge.

"It's a really exciting experiment," Kasparov says. "I'm looking forward to future challenges."
COPYRIGHT 1996 Science Service, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1996, Gale Group. All rights reserved. Gale Group is a Thomson Corporation Company.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:world chess champion Garry Kasparov beats chess computer Deep Blue
Author:Peterson, Ivars
Publication:Science News
Article Type:Brief Article
Date:Feb 24, 1996
Previous Article:Compound may cause wasting seen in cancer.
Next Article:Waterborne arsenic poses a cancer risk.

Related Articles
How to beat a chess champion.
Crowning a rookie chess champion.
Blitzing to win at computer chess.
Speeding to a chess championship.
Deep Thought for winning chess.
Computer chess: a masterful lesson.
Quick moves claim computer-chess title.
Computing a chess game's end.
The soul of a chess machine; lessons learned from a contest pitting man against computer.
Computer triumphs over human champion.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters