The idea of testing out part of Google’s developing artificial intelligence machinery came from Demis Hassabis, the Google DeepMind founder. Hassabis is a British artificial intelligence researcher, neuroscientist, computer game designer, entrepreneur, and world-class games player. DeepMind Technologies Limited is a British artificial intelligence company, purchased by Google’s parent company Alphabet in 2014. One of the test products is AlphaGo, which is ‘a narrow artificial intelligence’ computer program. The object of the platform is to play the board game Go.
One of the aims is to test whether AlphaGo can ‘learn’ how to beat the best Go human players in competition. For this AlphaGo has a special algorithm based on Monte Carlo tree search. This allows the platform to assess moves based on knowledge previously “learned” by machine learning. The learning is based on its artificial neural network (described as a deep learning method), where learning comes about through extensive training garnered from both human and computer Go competitions.
Go – the board game
Go is a 2,500 year old abstract strategy board game for two players, in which the aim is to surround more territory than the opponent (as players take turns placing stones on a 19-by-19 grid). Go is arguably more complex than chess. The game is said to possess more possibilities than the total number of atoms in the visible universe.
To see how far the artificial intelligence had progressed, a contest was set up with Ke Jie. Jie is a Chinese professional Go player, ranked number one in the world. However, with the contest, AlphaGo secured the victory after winning the second game in a three-part match.
The first match is shown in the following video:
The match was played at the Future of Go Summit in Wuzhen on May, 23 2017. Speaking with the BBC after the match, Ke Jie stated: “I’m a little bit sad, it’s a bit of a regret because I think I played pretty well.”
The aim of testing the artificial intelligence platform out has a serious and longer-term objective, which is to one day deploy such artificial intelligence into areas of medicine and science.