Google's AI subsidiary DeepMind has unveiled the latest version of its Go-playing software, AlphaGo Zero.
It is said to be even more powerful than its predecessor, which beat two of the world's best players, including Korea's Lee Sedol.
More importantly, AlphaGo Zero did not analyze any human moves and completely taught itself with a blank Go board and no data apart from the rules.
Within three days, it was good enough to beat the original program by 100 games to zero.
DeepMind CEO Demis Hassabis said his team is now working to apply the new AI technology into other real world problems.