It has been nearly 30 years since IBM developed the chess playing artificial intelligence system known as Deep Thought, which eventually took on then world-renowned chess champion Garry Kasparov. They played the first of two six-game chess matches in Philadelphia in 1996. Kasparov won this match. Deep Blue, however, won the second match in New York City the following year. This marked the first time a computer would defeat a world champion under grand tournament conditions.
More importantly, though, it marked the first real evidence of the power of artificial intelligence. The match was so monumental, in fact, it became the subject of the documentary film, The Man vs The Machine.
And in the last two decades since this match, artificial intelligence has vastly improved. In fact, “machine thinking” is one of the hottest areas of technological research and development right now. And at the forefront of this is none other than Google who has developed yet another machine to play a board game. This time, though, Google’s new AI system will take on the current world champion of the game ‘Go’.
The winner will take home a purse of one million dollars.
In order to understand why Google has chosen to take on a Go champion, it is necessary to first look at why a computer might be “better” than a human at chess. Chess is a surprisingly mathematical game. Every turn leads to new possibilities and the human mind calculate the outcomes to determine the best possible move. This is no different than a computer predicting who might win the Super Bowl based on statistics from the previous NFL season; except for the fact that humans can be unpredictable, and computers—at least, before Deep Blue—could not account for some of the randomness of human behavior.
But in the world of chess, the possible outcomes are limited. In the game of Go, however, any given match could yield a trillion or more possible moves. This, then, could be the true test of modern artificial intelligence.
And so, Google has developed its AlphaGo computer (probably in a nod to its parent company, Alphabet in a way similar to IBM naming their computer Deep Blue, after the company’s nickname, “Big Blue”). AlphaGo has already taken on and defeated the European Go champion in a five-game match. Now, of course, the machine will face the world champion, Lee Sedol, in another five-tournament.
The match will take place March 9-15 in South Korea.