Google's AlphaGo Just Beat The Chinese Grandmaster In A "God-Like" Victory

Ke Jie, left, and a human-interpreter of AlphaGo's moves, right. VCG via Getty Images

All hail our new robot overlords. China’s greatest Go player, 19-year-old Ke Jie, has just lost to an artificially intelligent player, heralding a new era in AI.

Yesterday, AlphaGo won its first of three matches against Ke Jie by half a point in Wuzchen, China, an extremely narrow margin. Ke Jie said he was shocked at the unexpected moves from the machine.

“AlphaGo is a completely different player,” he said after the game. “It is like a god of a Go player.”

In January this year, AlphaGo – built by Google's DeepMind lab – had already won 60 matches against the world’s top players, including Ke Jie. Thus, many had given the human little chance of winning before the match. You might remember it beat the Korean grandmaster Lee Sedol back in March last year. Now it’s defeated the Chinese grandmaster, too.

Lee Sedol was beaten 4-1 by the machine, which at the time was billed as a milestone in AI. Go is a unique, ancient game that has an incalculable number of permutations. Thus, it relies mostly on skill and intuition.

This latest match gives a good indication of how AlphaGo has progressed over time. Its play style has evolved dramatically from a year ago, with WIRED describing how AlphaGo had control of the board in just three and a half hours, whereas most had expected the game to last six hours. Ke Jie conceded defeat less than an hour later.

You can watch a replay of the match against Ke Jie with commentary above. DeepMind

Go, a Chinese board game stretching back 2,500 years, involves two players going head to head, placing black or white beads on a 19 x 19 grid to try and surround their opponent's bead. It’s somewhat similar to chess, but much more complex. In Go, there are 1 x 10170 possible configurations – a one with 170 zeroes after it – which is more than the number of atoms in the universe.

Chess was mastered by AI in 1997 thanks to the IBM supercomputer Deep Blue, which defeated Garry Kasparov in a landmark game. Go, however, presented a whole new challenge. It’s not only more complex, but also a vital part of China’s cultural history. So much so that China censored the match from being shown live, supposedly because a win for AlphaGo could damage national pride.

If you want to see the next two matches, you’ll be able to view them on Google’s DeepMind YouTube page – unless you’re in China. The battle might not be over yet, but all signs suggest it’s nearly time to surrender to the machines.

Comments

If you liked this story, you'll love these

This website uses cookies

This website uses cookies to improve user experience. By continuing to use our website you consent to all cookies in accordance with our cookie policy.