Gregor Eisenhorn
Well-Known Member
Well, as the title indicates shit just got real with AI.
AlphaGo (an AI) defeated Lee Se-Dol (a South Korean professional Go player of 9-dan rank [which is the highest possible btw]) today, in the first of 5 matches to test whether it can compete against a human player. Go (the name of the board game) is famous for having an extremely high amount of possible board states, thus being much more difficult than, say, chess for a machine to "play". Overall, the game was very close and ended in a very late-game (<10 min on AlphaGo's clock) concession from Lee. A lot of people think his next game will be much easier (having seen the software in action/gauged its decision-making skills), however, past trials have shown that if a program can win ONCE against a Pro, it usually takes the series in a 5-0 sweep. You can watch the upcoming games on DeepMind's YouTube channel.
AlphaGo (an AI) defeated Lee Se-Dol (a South Korean professional Go player of 9-dan rank [which is the highest possible btw]) today, in the first of 5 matches to test whether it can compete against a human player. Go (the name of the board game) is famous for having an extremely high amount of possible board states, thus being much more difficult than, say, chess for a machine to "play". Overall, the game was very close and ended in a very late-game (<10 min on AlphaGo's clock) concession from Lee. A lot of people think his next game will be much easier (having seen the software in action/gauged its decision-making skills), however, past trials have shown that if a program can win ONCE against a Pro, it usually takes the series in a 5-0 sweep. You can watch the upcoming games on DeepMind's YouTube channel.