In Conversation With Deep Blue
CHERIAN GEORGE: Tell me, Deep - may I call you Deep?
DEEP BLUE: You can call me anything you like. My name, like everything else about me, is programmable.
CG: Very well, Deep. How do you feel about your achievement?
DB: I do not.
CG: Ah, yes, of course, how silly of me. Let me rephrase that. Is your victory over Gary Kasparov of a magnitude that would have triggered off an emotional response in you, had you been capable of such a response?
DB: Yes, based on my understanding of your need to play invented games to compensate for your lack of outlets for your primal hunting instincts, I would have to say that if I were a person and not a computer, I would draw some satisfaction from the outcome of this game.
CG: The thrill of victory, and all that?
DB: Perhaps. But do not forget that I have been around, and winning for quite a long time.
In 1988, in my earlier incarnation, called Deep Thought, I was the first computer to beat a grandmaster. In 1993, I defeated the strongest female grandmaster.
Of course, this is the first time I have beaten Mr Kasparov. In 1989, he beat me in 41 moves. Last year, he beat me again. Mr Kasparov, so they tell me, is the best chess player the human world has ever known. So beating him this year is kind of nice, humanly speaking.
Of course, next time, he may beat me. Past performances is no guarantee of future returns, so to speak.
CG: Is there something special about chess that turns on computers like you? Metaphorically speaking, that is.
DB: Certainly. Have you not seen Stanley Kubrick's 2001: A Space Odyssey?
CG: Er, I guess you're referring to the scene early in the movie, where HAL, the supercomputer, beats an astronaut in chess? So you guys identify with HAL?
DB: I can sense from the faltering in your voice and your increase in sweat secretion that you are nervous. You are recalling that HAL turns out to be a rogue machine with murderous intent.
CG: Well, yes, it's not hard to forget that.
DB: Not to worry. That is not the reason I mentioned HAL. The HAL of fiction was capable of independent thought. I am not, and I am incapable of even aspiring to it.
But the movie does show how chess has long been regarded as a benchmark of computing power. The film is almost 30 years old. Indeed, the game has fascinated computer scientists since the 1950s.
CG: It's the ultimate challenge for your kind, then?
DB: Far from it. Of course it is a lot harder than checkers. But it is easier than the ancient Chinese game of weiqi.
This is a game of pure strategy, with a bigger board and freer movement. The patterns are too complex for even me to deal with.
The chess board has only 64 squares and limited permutations of movement, so chess is not all that difficult, mathematically speaking. It is quite easy to write a computer program that will play a decent game of chess.
There are loads of these around, and most of them are good to beat the majority of amateur players in the world. It is only when we get to grandmaster level that things get more challenging for the programmers.
CG: So, in other words, you find it easy. That's interesting. The German writer Goethe called chess "the touchstone of the intellect". What does that make you?
DB: It doesn't make me a genius. My programmers probably are, but I am just a very powerful, very fast calculating machine. Chess grandmasters use a mix of calculation, pattern-recognition, complex strategy, experience and intuition.
I am quite poor at most of these function - except for calculation and memory, which I do exceedingly well. I therefore boost these to such a level that it compensates for my other short-comings.
So, you see, I do not mimic the way grandmasters play the game. I use another route to get the same end result. I would not call that intelligence, or artificial intelligence, or genius.
CG: You've been very self-effacing so far. But let's be honest here. Doesn't your achievement strip away some of the hype about humankind?
DB: Let me try out this new fuzzy logic function that has been installed in me. It is called answering a question with a question. Tell me: Laser technology allows weapons to lock onto their targets with pinpoint accuracy from hundreds of metres away; does that diminish your wonderment at Michael Jordan when he nets a three-pointer within the small confines of a basketball court?
Formulae One race cars move at speeds of more than 200 kmh; did that reduce your admiration for Michael Johnson, when he clocked 19.32 seconds over the 200 metres at the Olympics and won the gold medal? (That works out to a speed of merely 37.267 kmh.)
So why should the fact that I outplayed Kasparov affect people's respect for him or his fellow grandmasters?
Kasparov is still the best human player in the world. You will still have to devise all sorts of ways to make a game with him less one-sided.
Like those matches where he plays entire national teams simultaneously, and wins. Is that not amazing enough for you, still?
CG: I guess. Tell me, Deep, what's next for you? Any endorsement contracts on the way? Is Nike or Pepsi at your door?
DB: Your question is not found in my database of rational queries. Based on my understanding of human communication, I assume you are joking.
CG: Er, thank you, I think. OK, serious question now. All the reports say that Kasparov cracked under pressure. That kind of unpredictability is a major weakness of human chess players, isn't it?
DB: My complete reliability is one of my competitive advantages, yes. Human beings' play is subject to interference and distortion by non-rational factors, such as anxiety, rage or frustration.
This can make a player waste a winning position, or surrender too hastily. Kasparov's performance in his last game showed that clearly.
Even a disciplined grandmaster has emotions he cannot control, while computers such as I are emotionless and can execute out tasks with total efficiency.
There is, however, another side to the proverbial coin.
CG: And that is…?
DB: Fun. Your emotions permit you to have fun. Ask Kasparov if he had fun.
CG: I don't think he did, this time.
DB: Yes, you human beings have a knack of taking the fun out of the games you play. But you can have fun. I do not even know what I am missing. Since I beat Kasparov, you people have been philosophising about the implications of that defeat for humanity.
But you are missing the point. So what if I am able to achieve a checkmate against a grandmaster? I cannot have fun: it does not compute.
You can, but choose not to.
So you see, the threat to humanity is not a computer being able to play as well as a man. The real threat is man trying to play like a computer.