Google's AlphaGo A.I. beats world's number one in ancient game of Go

Demis Hassabis, co-founder of Google's artificial intelligence (AI) startup DeepMind.
Jeon Heon-Kyun | Getty Images

Wuzhen, China -- Computer 1, human 0. The rise of the machine was fully evident Tuesday when a Google computer beat the world's number one player of the ancient Chinese board game Go, signifying a major breakthrough in artificial intelligence.

, lost his first game against Google DeepMind computer program AlphaGo in Wuzhen. With two remaining matches, Ke, 19, could still obliterate AlphaGo and take home $1.5 million in prize money, maintaining his status as the absolute best. But if he loses, that would solidify machine domination in one of the world's most complex games.

Despite defeat in the first match, "this is the greatest honor in a lifetime," Ke said of playing against AlphaGo.

Teaching computers to master Go has long been considered a holy grail for artificial intelligence scientists — there are more possible configurations of the board than there are atoms in the universe. Before this week, AlphaGo had already clocked many victories against top-ranked masters, a significant advancement that happened far sooner than experts expected.

"I've always been optimistic about how far we can go with AI," Dave Silver, lead programmer of AlphaGo at Google DeepMind, told CNBC. "But it happened faster than I think even I would have anticipated – we were able to really discover some beautiful and elegant algorithms … leading to some nice results against the world's strongest players."

AlphaGo: What's next for A.I.?
AlphaGo: What's next for A.I.?

This week's games between AlphaGo and Ke are being live streamed by Google online, but Chinese censors already block access to Google and its services, such as Gmail and YouTube, in the country.

Chinese state media outlets were also restricted by authorities from live broadcasting the game, and issued coverage guidelines ahead of the match, according to sources familiar with the matter, who declined to be identified on the record because they weren't authorized to speak in an official capacity.

It was unclear why, given that Go is a Chinese game, and Ke himself is Chinese, though it could be linked to the broader ban on Google in China. A Google spokesperson declined to comment.

The game, Go, originated thousands of years ago in China, and has two players taking turns placing black and white stones on a square board of 19 lines by 19 lines. The object is to take territorial control of the board by surrounding the opponent. Games can go on for hours, and playing requires immense mental stamina, intuition, and strategy.

AlphaGo was trained by studying how human experts have played in the past, Silver said. The program runs a very deep search, sometimes gaming out 50 moves into the future, and uses "this tree of possibilities to work out the best path that will lead to the best possible outcome," Silver said.

This is the kind of technology that Google DeepMind wants to apply elsewhere, to help humans figure out what present decision will lead to the best potential results.

DeepMind's CEO on machine learning
DeepMind's CEO on machine learning

"We're most excited about applying AI to science and medicine … using AI as a tool to improve scientific breakthroughs," Demis Hassabis, CEO of Google DeepMind, told CNBC. "And also to improve healthcare and maybe eventually help with finding cures for diseases or medical diagnostics."

But scary machines aren't going to replace humans entirely just yet. As it stands, computer systems can't replicate things like human emotions and imagination. And Hassabis says there's a long way to go for machines to process higher cognitive functions, like memory, planning, and abstract reasoning.

Still, the technology has come a long way since the 1990s, when software programs first got smart enough to play classic board games, like backgammon. Things peaked with a historic victory of IBM's Deep Blue computer over world chess champion Gary Kasparov in 1997.

Though not a serious Go player -- Hassabis made his name in his early teens as a world chess champion -- even he hasn't tried to play against AlphaGo.

"There would be no point in me playing AlphaGo," he said. "It's way too strong for me."