Gigaom AI Minute – March 11

Transcript

Today is March 11. We're looking back at the 2016 Go tournament between Lee Sedol and AlphaGo. As of yesterday, March 10, Lee Sedol was down 2-0 in a five-match tournament. He would need three straight victories to win, but in the minds of many the goal had been changed to simply winning once.

March 11 was a day off from play. Lee is said to have spent the day resting and talking to other Go players. One can imagine that even among adversaries, this situation might make one rise above normal considerations of sport and band together to defeat the machine. It was also a chance for everyone to reflect. There seemed to be a profound sadness, a vague sense that somehow humanity was diminished by the events of the prior two days.

Is it sad when an AI defeats a human in a new field, whether it be chess or Go or poker or any other uniquely human endeavor? For many people, the answer is an unqualified yes. It's sort of like a 'Rocky' movie in which Rocky fights an alien from another planet or King Kong or Godzilla. How could someone not root for Rocky?

But I see things differently. I don't think it's sad at all. The thing to remember is that we made AlphaGo. It's still us playing the game of Go. We're simply using a new tool to play it better. Does anyone ever look at Michelangelo's statue of David and find it sad that he used chisels to sculpt it? Do you admire the giant skyscraper any less because cranes lifted the beams into place? One can argue that the situation is different because AlphaGo isn't helping us do something, it's doing it without us. But that isn't quite right. Every time AlphaGo would play a new opponent, the programmers would study the games and figure out how to make the AI better. It still comes back to us.

And if it's true that Go is a beautiful game when played well, then doesn't AlphaGo make it even more beautiful?