Add “becoming world number one Go player” to the list of achievements technology has just made it pointless to pursue.
Proving that its earlier success was not a fluke, Google's AlphaGo AI has once again beaten the world's number one ranked human player of the age-old game.
After the first round of a three part match between the DeepMind-powered AI and Ke Jie, the champion human player, the AI was able to defeat its human rival despite Ke Jie employing tricks he'd learnt from watching the AI play.
19-year old Chinese Go guru Ke follows in the esteemed footsteps of Go-playing legend Lee Se-dol, who lost to AlphaGo 4-1 in a similar tournament a little over a year ago.
Going, going, gone
It was a close-fought match, but Ke Jie appeared to technically be on the backfoot from early on. What's particularly interesting was AlphaGo's playstyle – it scraped the win, but in such a way that suggested its margin of victory was never a concern. It didn't aim to trounce Ke Jie, but simply to beat him by whatever means.
“I think it was a really wonderful game,” DeepMind CEO and co-founder Demis Hassabis said after the match. “Huge respect to Ke Jie for playing such a great game and pushing AlphaGo to its limits.”
The match is part of the Future Go Summit, being hosted by Google in Wuzen China. Rematches will take place on Thursday and Saturday before an ultimate winner will crowned.
But there's no rest for AlphaGo – it's also going to be acting as a teammate to two Chinese players facing off against each other, as well as taking part in another match against five pro-level human opponents at once.
- AI and automation will take our jobs – but where and when?