In May 1997, an IBM supercomputer known as Deep Blue beat then chess world champion Garry Kasparov, who had once bragged he would never lose to a machine.
Kasparov and other chess masters blamed the defeat on a single move made by the IBM machine. Either at the end of the first game or the beginning of the second, depending on who's telling the story, the computer made a sacrifice that seemed to hint at its long-term strategy.
Kasparov and many others thought the move was too sophisticated for a computer, suggesting there had been some sort of human intervention during the game. "It was an incredibly refined move, of defending while ahead to cut out any hint of countermoves," grandmaster Yasser Seirawan told Wired in 2001, "and it sent Garry into a tizzy."
Fifteen years later, one of Big Blue's designers says the move was the result of a bug in Deep Blue's software.
The revelation was published in a book by statistician and New York Times journalist Nate Silver titled The Signal and the Noise — and promptly highlighted by Ezra Klein of the Washington Post.
For his book, Silver interviewed Murray Campbell, one of the three IBM computer scientists who designed Deep Blue, and Murray told him that the machine was unable to select a move and simply picked one at random.
At the time, Deep Blue versus Kasparov was hailed as a seminal moment in the history of computer science — and lamented as a humiliating defeat for the human intellect. But it may have just been a lesson that as humans, we tend to blow things way out of proportion.
Many chess masters have long claimed that Kasparov was at a significant disadvantage during the match. Deep Blue's designers had the opportunity to tweak Deep Blue's programming between matches to adapt to Kasparov's style and strategy. They also had access to the full history of his previous public matches.
Kasparov had no similar record of Big Blue's performance. Because the machine had been heavily modified since he had last played it, he was essentially going in blind. That strange move was chalked up to these advantages.
The IBM team did tweak the algorithms between games, but part of what they were doing was fixing the bug that resulted in that unexpected move. The machine made a mistake, then they made sure it wouldn't do it again. The irony is that the move had messed with Kasporav's mind, and there was no one to fix this bug.
"Kasparov had concluded that the counterintuitive play must be a sign of superior intelligence," Campbell told Silver. "He had never considered that it was simply a bug."
It's tempting to think there's a lesson here about human nature. After all, a human mistake in the development of the software led to the machine's victory. It's sort of reassuring to think that a human flaw is actually what made Deep Blue successful. But it's not clear that things would have turned out all that differently had that bug never surfaced.
Years after the final Deep Blue match, both Kasparov and Vladimir Kramnik, his successor as world chess champion, played against various versions of Deep Blue's successor Fritz. But in these matches, no code modifications were permitted between games. Kramnik even had the chance to play against the software in advance of the matches, and had the right to adjourn a game until the next if it went past 56 moves.
The results aren't that encouraging for humans.
Kasparov's match against X3D Frintz in 2003 ended in a draw. So did Kramnik's first match against Fritz in 2002. And Kramnik lost to Fritz due to a blunder in 2006.
These weren't decisive victories for the machines, but the humans still couldn't win. Even though humans can conceive of strategies to counteract the computation advantage of computers, we get tired, make blunders, and suffer from anxiety. Machines never get tired or flustered.
But the relationship between chess players and computers is actually more symbiotic than adversarial. Today's chess masters use computers extensively as learning aids.
That said, today's computers make Deep Blue look puny. Maybe it's time for a rematch.
