Google's Artificial Intelligence (AI) program has beaten the European ‘Go’ champion in a competition. The news was announced by DeepMind, the company’s AI arm, which has headquarters in London, United Kingdom. Analysts are calling this a major milestone for AI as ‘Go’ is much more difficult to play than chess (which computers have been playing for many years). While typically there are 20 moves available to chess players for any Chess move, go offers players around 200 possible options.
Google’s AlphaGo software analyzed 30 million moves from previous games in preparation for its win. The final score was AI ‘five’, human champion ‘nil’. Google’s AI program has surpassed that of Facebook which has yet to beat a human ‘Go’ player. The scientific journal ‘Nature’ has released a paper outlining the AI techniques used to win the game.
‘Go’ is a traditional Chinese game which is thousands of years old and also played in Japan and Korea. It is a board game and has two players. The objective of the game is to surround territory on a board by overturning black-and-white stones - the player with the most territory wins the game.