Magazine article AI Magazine

Deus Ex Machina-A Higher Creative Species in the Game of Chess

Magazine article AI Magazine

Deus Ex Machina-A Higher Creative Species in the Game of Chess

Article excerpt

Computers and human beings play chess differently. The basic paradigm that computer programs employ is known as "search and evaluate." Their static evaluation is arguably more primitive than the perceptual one of humans. Yet the intelligence emerging from them is phenomenal. A human spectator cannot tell the difference between a brilliant computer game and one played by Kasparov. Chess played by today's machines looks extraordinary, full of imagination and creativity. Such elements may be the reason that computers are superior to humans in the sport of kings, at least for the moment.

Not surprisingly, the superiority of computers over humans in chess provokes antagonism. The frustrated critics often revert to the analogy of a competition between a racing car and a human being, which by now has become a cliche.

In 2007, Amir Ban (my partner) and I flew to Elista, capital of Kalmkya, to take part in a chess match dubbed "The Ultimate Computer Chess Challenge." Our program, Deep Junior, then the world computer chess champion, was to play a match against Deep Fritz, which a few months earlier had defeated Vladimir Kramnik, the world's human chess champion.

Parallel to the computers' match, at the same venue, some of the best human chess players assembled to play in the semifinal tournament for the world title. FIDE, the international organizing body, had asked top commentators to annotate the games in both competitions. Among the commentators was Boris Spasky, the famous former world champion.

Some of the games between Deep Junior and Deep Fritz were exciting: in one of them, Deep Junior was able to discover an important theoretical novelty in the Sicilian defense. After sacrificing no fewer than four pawns, it managed to find a winning maneuver that was not known beforehand. Disappointingly, the commentators shied away from the computer games, refraining to analyze them.


Mig Greengard, an influential chess commentator, tried to explain why humans avoid computer games. According to him, spectators prefer to watch how chess icons lose by committing terrible blunders of the kind that do not occur any more in computer chess. Perhaps this provides them with consolation for their own imperfections.

Perhaps Greengard's comment agrees with Grandmaster Robert Huebner's view that "we play chess because we do not know how to play it"; in other words, mistakes are just part of the fun of playing. However by ignoring computer games, humans miss the great progress that has been made and the beauty of their play--just as if one chose to ignore a beautiful picture merely because it was painted by a computer.

Some of the people who take an interest in computer chess admit computer superiority but cast doubt over the methods employed to achieve it. They revive the term brute force in reference to the overwhelming computer search mechanism, pointing out that mortal players reach comparable results looking at many fewer positions. While indeed human abilities really are amazing, beneath their hood hide millions of neurons probably performing primitive calculations as well. To quote Marvin Minsky, "We are not just simple machines, we are wonderful ones."

This article is about how roles have changed: humans play chess like machines, and machines play chess the way humans used to play.

Computer Intuition and Concrete Play

It is conceivable that the computer's "chess neuron" (a role attributed to its static evaluation function) is superior to the human one because it is designed to perform a limited yet special task. Either way, the net result in both cases is the emergence of a creative thinking process. The spectator witnesses the beautiful expression of the machine's inner self and should not be bothered by all those neurons. In this sense, computer chess programs pass the Turing test big time.

Another interesting analogy exists in a comparison between the intelligent search process that occurs in clever computer chess machines and swarm intelligence (SI). …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.