Gigaom AI Minute – March 10

:: ::

In this episode, Byron talks about AI's creativity when it played Lee Sedol in chess.

Transcript

On this date, March 10, in 2016, the second game was played in the match between Lee Sedol and AlphaGo.

As I wrote yesterday, the world had been shocked by the outcome of Game 1, in which AlphaGo forced Lee to resign. This unexpected victory for the computer brought even more media attention to the scene than before.

Around Move 36, Lee left to smoke a cigarette outdoors as was his habit when he was feeling a lot of pressure. While he was gone, AlphaGo made the legendary Move 37.

Move 37 took everyone by surprise. Live commentators looked at it and scratched their heads, unable to ascertain right away if it was a brilliant move or a computer bug. The AlphaGo team scrambled to figure out what was going on, and when they queried AlphaGo with their diagnostic tools, they learned that, based on the games that AlphaGo had been trained on, it calculated that there was only a 1-in-10,000 chance that a human would have played that move.

Lee returned to the board and saw Move 37. He just studied it awhile, and eventually saw the brilliance of the move, which he later called "creative and beautiful." Move 37, simply put, lost the game for Lee Sedol, and it was a move that no human player would have made. Immediately after the game, commentators began referring to AlphaGo's "creativity." Lee would later say that while Game 1 impressed him, Game 2 blew him away.

Was this a creative move? The concept of creativity is so loosely defined even in our minds that it is hard to say whether this is a move that was creative or simply one that looked creative, or if there’s a meaningful difference between those two statements. Clearly, though, this was a watershed moment in the history of computers, and perhaps of humanity.

Interested in sponsoring one of our podcasts? Have a suggestion for a great guest? Please contact us and let us know.