Google AI beats Go world champion again to complete historic 4-1 series victory


The battle between Google’s artificial intelligence and Go world champion Lee Sedol concluded today after the former (AlphaGo) triumphed to win the five-game series 4-1.

AlphaGo, created by London-based DeepMind — which was acquired by Google for around $500 million in 2014 — found itself in a little trouble during today’s final match in Seoul, but it managed to turn it around and take victory, DeepMind CEO Demis Hassabis said on Twitter.

That AlphaGo was able to recover from an error against one of the best players on the planet is a real testament to the its ability, and the development of AI to date. More widely, its series victory marks a historic landmark for the development of artificial intelligence. Go is an east Asian game that is famous for its deep complexity, boundless possibility of moves and strategic thinking. For AI researchers like Hassabis, it is the ultimate test of their creations.

Pundits were divided on AlphaGo’s potential to win in the lead in to the matches. The AI beat the European Go champion last year, but Lee — a legend in the game — was seen as a far stiffer challenge.

Unlike previously AI victories — such as Deep Blue’s defeat of chess grandmaster Garry Kasparov in 1997, or IBM Watson’s Jeopardy! triumph in 2011 — DeepMind programmed AlphaGo to be capable of teaching itself, not just carrying out a set of fixed moves or activities. Indeed, the fact that Lee chalked up one match victory and pushed the machine hard shows that it is not impenetrable and is also capable of making mistakes or being outwitted. In other words, there’s still scope for further progress and development despite this incredible achievement.

“The winner here, no matter what happens, is humanity,” Alphabet chairman Eric Schmidt said prior to the first match, and that certainly seems to ring true now the challenge — and $1 million purse riding on it — has been decided.

You can look through the games via the DeepMind YouTube channel.

Source link

Comments are closed.