Google DeepMind is going to pitch its AI against the world's best Go player and broadcast it live on YouTube
The announcement, made on Twitter last night, comes less than a week after DeepMind said it had beaten the best Go player in Europe, Fan Hui.
The match, which will involve a $1 million (£680,000) wager, will be played in Seoul, South Korea next month and live-streamed over YouTube.
Go, the eastern version of chess, is a board game for two players that is thought to have originated in China more than 2,500 years ago and has trillions of possible moves.
While computers can beat humans at chess and other games, looking ahead in Go is more difficult for them because the number of possible moves is far greater.
DeepMind chief Demis Hassabis announced the news on his Twitter account:
The London startup, which was acquired by Google for £400 million in 2014, was in a race with Facebook to develop an AI that could defeat a professional Go player. DeepMind can now do that but Facebook can only beat amateur players.
Hassabis is hailed as something of a genius, having gained chess master status at the age of 13 and a double first in computer science from Cambridge University.
However, AI experts have questioned how significant Google's AlphaGo algorithm really is. Rodolfo Rosini, cofounder of London AI startup weave.ai, said the company is better at marketing than AI.