Welcome to another edition of our Famous Bugs series! We hope you get a serious laugh out of other people’s pains that came before us. In this post, we’Il explain the fascinating history of the IBM Deep Blue bug.
In May of 1997, IBM’s chess-playing computer Deep Blue challenged the World Champion of Chess, Russian Grandmaster Garry Kasparov, to a single game that would make history. Having once bragged that he would never lose to a machine, Kasparov did just that.
It’s the backstory that intrigues psychologists, psychiatrists, artificial intelligence engineers, and computer scientists to this day. Kasparov along with other chess masters blame the defeat on one single move made by IBM’s Deep Blue. The computer apparently made a sacrifice that seemed to hint at its long-term strategy. The computer chose not to capture an exposed pawn in favor of another route — an odd sacrifice that back then seemed far beyond the strategic foresight of a computer.
The chess grandmaster was in complete disarray after Deep Blue made this move and believed the move was too sophisticated for a computer to make on its own. He went so far as to suggest some sort of human intervention took place during the game.
Grandmaster Yasser Seirawan famously stated that “it was an incredibly refined move, of defending while ahead to cut out any hint of countermoves”.
It was only FIFTEEN years later when one of IBM’s designers revealed that the move was the result of a bug in the Deep Blue software.
This IBM Deep Blue bug revelation only came to light when it was published by Nate Silver, a renowned statistician and journalist for the New York Times, in his book The Signal and the Noise. In an interview with Murray Campbell, one of the three IBM computer scientists who designed Deep Blue, Murray explained that the machine simply picked a move at random due to its inability to select a move in that specific moment.
Deep Blue’s victory was regarded as one of the great breakthroughs in computer science at the time. It seemed to be a turning point in the potential supremacy of artificial intelligence over human intellect.
On the other hand, many chess masters came to Kasparov’s defense saying that he was at a complete disadvantage. This was due to the fact that Deep Blue’s designers had the opportunity to make tweaks and adjustments to the software between matches according to Kasparov’s style of play. The argument for Kasparov is that there was no equal footing since he had no access to Deep Blue’s performance record and was hence going in blind.
The IBM Deep Blue team did make tweaks between games so that any mistakes would not to be repeated. The only problem was that this random move — a bug — sent Kasparov into a complete frenzy, and there was no one to fix this bug. He immediately thought it was a move of superior intelligence and not just a random software bug.
We see a couple of takeaways from this story. First, don’t panic in the face of uncertainty like Kasparov — some things just happen. Second, for developers, it turns out that not all bugs are harmful bugs.
STAY TUNED FOR MORE FROM OUR FAMOUS BUGS SERIES AND LET US KNOW IF ANY BUGS ARE GETTING THE BEST OF YOU! 😉