The Telegraph My details My newsletters Logout Upgrade to Premium My details My newsletters Logout The Telegraph Google's machine beats master at ancient game of Go in historic AI breakthrough 27 January 2016 • 6:00pm A computer program developed by Google has beaten a professional player at the ancient strategy game of Go, a potentially-historic breakthrough in the quest to develop artificial intelligence capable of matching humans.
 DeepMind, a British AI company owned by the internet giant, has developed software that repeatedly beat the European champion of the fiendishly-difficult board game, an advance that has evaded researchers for years and was believed to be a decade away.
--
 Instead, DeepMind developed a program that learned to play Go in a similar way to humans, by playing repeated games and figuring out the moves that are most likely to lead to success.
 Researchers believe these “deep learning” techniques could be used to build artificial intelligence capable of surpassing humans in many areas.
 The AlphaGo software was able to scan the gaming patterns of world-class players as well as playing millions of games against itself, learning by trial and error.
--
 Until now, the most advanced Go programs were only able to match up to amateur players.
 Gary Kasparov losing to IBM's Deep Blue at chess in 1996 Gary Kasparov was defeated by IBM's Deep Blue at chess in 1997 Credit: EPA The breakthrough, published in the scientific journal Nature, is one of the “long-standing grand challenges of artificial intelligence”, DeepMind’s Demis Hassabis said.
 “This has been the holy grail since Deep Blue beat Kasparov at chess, and it’s held out for over 20 years.
--
 The same techniques used to teach AlphaGo to play Go could be used to develop digital assistants that will automate parts of our daily lives, diagnose medical conditions faster than human doctors and help solve major scientific challenges such as modelling climate change and curing diseases, Mr Hassabis said.
 Facebook, which like Google is racing to develop artificial intelligence, has also been attempting to build a computer that can beat professional Go players.
 The ancient Chinese game of Go is one of the last games where the best human players can still beat the best artificial.