01  02  03  04  05  


06  

Inside Chess Online

The chess scene in 2001 is just one example of the genius behind Clarke and Kubrick's screenplay. Although the game between HAL and astronaut Frank Poole is shown for only about thirty seconds, it conveys a great deal of information about HAL and the relationship between Frank and HAL. The fact that HAL can beat Frank at one of the world's oldest and most difficult games is clearly intended to establish HAL as an intelligent entity. But is this a correct conclusion? Does a machine need to be intelligent to play chess?

The question of whether HAL's chess ability demonstrates intelligence boils down to a question of how HAL plays chess. If, on the one hand, HAL plays chess in the "human style" -- employing explicit reasoning about move choices and large amounts of chess knowledge -- the computer can be said to demonstrate some aspects of intelligence. If, on the other hand, HAL plays chess in the computer style -- that is, if HAL uses his computational power to carry out brute-force searches through millions or billions of possible alternatives, using relatively little knowledge or reasoning capabilities -- then HAL's chess play is not a sign of intelligence.

This chapter attempts to resolve this question by examining in detail how HAL plays chess and by comparing HAL with Deep Blue, the world's current premier chess computer. I and my colleagues, Feng-hsiung Hsu and A. Joseph Hoane, Jr., developed Deep Blue at IBM's T. J. Watson Research Labs. It was the first machine in history to beat the human world champion, Garry Kasparov, in a regulation chess game. The chapter also examines the strengths and weaknesses of computer-style chess by looking at some of the games between Kasparov and Deep Blue. Finally, we discover that HAL's first error occurred in the chess game with Frank.

Before we analyze how HAL plays chess, we need to put his game with Frank into perspective by understanding the history of man-machine chess matches. What is the significance of a machine beating a human at a game like chess?


Frank versus HAL; Man versus Machine

HAL claims to be "foolproof and incapable of error." But, as we witness only one isolated game between Frank and HAL, how do we really know that HAL plays well? The answer can be determined, not so much by the game itself but by Frank's reaction to it.

Poole: Umm ... anyway, Queen takes pawn.

HAL: Bishop takes Knight's Pawn.

Poole: Lovely move. Er .. Rook to King One

HAL: I'm sorry, Frank. I think you missed it. Queen to Bishop Three. Bishop takes Queen. Knight takes Bishop. Mate.

Poole: Ah ... Yeah, looks like you're right. I resign.

HAL: Thank you for an enjoyable game.

Poole: Yeah. Thank you.

Having personally witnessed scores of amateur chess players lose to computers, I found Frank's reaction to losing to HAL extremely realistic. After HAL announces mate, Frank's pause is brief. This brevity is significant, because it demonstrates that Frank assumes HAL is right. He trusts that HAL has the details of the checkmate correct and does not take the time to confirm them for himself. Instead, Frank resigns immediately. Moreover, it is obvious from his tone of voice -- or perhaps I should say from his complete lack of tone -- that he never expected to win. In fact, Frank would have been utterly stunned if HAL had lost. No, playing chess with HAL is simply a way for Frank to pass the time on the eighteen-month journey to Jupiter. (As HAL is running virtually every aspect of the ship, there is little for the two, nonhibernating astronauts to do.) It is also clear from the dialogue, as well as from Frank's body language, that this is not a game between two competitors but one between two conscious entities -- one of whom is vastly superior in intelligence to the other.

Clearly, Frank does not feel bad about losing to a computer, any more than a sprinter would feel bad about being outrun by a race car. Nor do we, the viewers, feel particularly sorry for Frank's loss. We don't mind HAL winning, because at this stage in the film we like HAL. The human relationship with chess computers hasn't always been so amicable though.

In many recent human-machine matches, the mood has been decidedly pro-human and anti-computer. In the first encounter between the human world champion (Kasparov) and the computer world champion (Deep Thought, Deep Blue's predecessor) in 1989, there was definite hostility toward the computer. When Kasparov pulled out the victory, the audience breathed an audible sigh of relief. In gratitude for "saving human pride," onlookers gave Kasparov a standing ovation.

Kasparov couldn't, however, save humanity's pride indefinitely. In 1995 he lost a game of speed chess to a computer program called GENIUS3. Burying his head between his hands, Kasparov could not hide his despair; he stormed off the stage, shaking his head in disbelief. The loss, reported by newspapers and magazines around the globe, shocked the multitude of those -- players and nonplayers alike -- who believed that the strongest player in the history of the game would never suffer defeat at the hands of "a silicon monster." Although Kasparov was badly shaken by this upset, it was, after all, only speed chess -- a game in which decisions are made within severe time constraints. (Speed chess allows each player only twenty-five minutes for the entire game, whereas players in regulation chess each have two hours to complete forty moves.)