the Division of DeepMind Google has developed a system AlphaGo that are perfectly cope with the Chinese Board game of go. As stated by Engadget, the possible combinations of moves in there are more than atoms in the universe. This statement is difficult to verify, but the program proved today it has no equal in this Board game.

Read more...

a Division of DeepMind Google has developed a system AlphaGo that are perfectly cope with the Chinese Board game of go. As stated by Engadget, the possible combinations of moves in there are more than atoms in the universe. This statement is difficult to verify, but the program proved today it has no equal in this Board game.

Instead of calculate all the possible moves, the team AlphaGo taught his ward 30 million movements, which are used by professional players. Then the program played against itself, practiced and created my own strategy by trial and error. Only then AlphaGo clashed with three-time European champion, a native of China whose name is Fang Hui.

the result is flawless victory remained with the program Google, which was supported by the neural network. As experts say, "this is only the first step on the way to the solution of artificial intelligence". In March AlphaGo have to fight with the world champion Lee Segalem.

See also:

the
• Chinese programmers as a bonus gave rubber womens and Slippers
the
• In Minsk of the "pros" earn more than in Moscow and Kiev;
the
• News sites are slowing down the spread of epidemics