The world's best player of Go (the world's most complicated board game) was roundly defeated by a Google program called AlphaGo in the first game of their three-game match this week, per the NYT.
Why it's so complicated: Go has an incalculable number of possible moves, meaning that a supercomputer can't simply crunch the numbers. Instead, AlphaGo uses artificial intelligence to learn new techniques each time it plays — even learning more by playing against itself. Because Go has been played for thousands of years, AlphaGo's unorthodox moves can ruffle grandmasters like Ke Jie, the world's best player.
What's next: Ke defeated AlphaGo last year, but said now that it's "improving too fast," calling it "a different player this year compared to last year." Google hopes to transfer AlphaGo's exponential learning success to other applications, like scientific research and medical diagnoses.