For AI researchers, Go is equally inspired. Chess fell to the computers in 1997, when Garry Kasparov lost an event to Deep Blue, an IBM computer. But until Mr. Lee’s defeat, Go’s complexity had made it immune to the march of machinery. AlphaGo’s success was an eye-catching presentation of the power of a type of AI called machine learning, which aims to get machines to teach complicated tasks to themselves.
AlphaGo acquired to play Go by studying thousands of games within expert human opponents, plucking rules and strategies from those plays and then refining them in millions of more matches which the details played against itself. That was sufficient to make it stronger than any human player. But researchers at DeepMind, the firm that built AlphaGo, we’re sure that they could improve it. In a paper just written in Nature, they have revealed the latest version, dubbed AlphaGo Zero. It is much better at the game, learns to play much more swiftly and requires far less computing hardware to do well. Most significant, though, unlike the original version, AlphaGo Zero has accomplished to teach itself the game without recourse to human experts at all.
Like all the best games, Go is straightforward to learn but hard to master. Two players, Black and White, take turns arranging stones on the intersections of a board consisting of 19 vertical lines and 19 horizontal ones. The aim is to dominate more territory than your opponent. Stones that are enclosed by opponents are removed from the board. Players carry on until neither wishes to continue. Each then adds a number of his stones on the board to a number of empty grid intersections he has surrounded. The larger total is the winner.
The difficulty arises from the sheer number of possible moves. A 19×19 board offers 361 various places on which Black can put the initial stone. White then has 360 choices in response, and so on. The total number of legal board compositions is in the order of 10170, a number so large it defies any mechanical analogy.
Take your time to comment on this article.
