The DeepMind artificial intelligence division of Google has effectively demonstrated their self-teaching AI’s superiority over “Stockfish 8,” the world’s foremost digital chessmaster.
University of Oxford Professor Michael Wooldridge told the BBC that this triumph is “the latest in a series of dazzling results that DeepMind has produced.” He said that AlphaGo Zero’s “trajectory” seems to be the ability to “solve a problem, and then demonstrate it can really ramp up performance — and that’s very impressive.”
The team has already conquered an arguably far more complex task in the ancient game of Go, for which the AI itself is named. It has also taught itself to play simple video games like Pong and Space Invaders, and is currently being refined toward becoming the ultimate Starcraft competitor.
While Google refuses to comment on this performance before it is officially published in a research journal, Cornell University has reported that AlphaGo Zero was able to beat Stockfish into a standstill only four hours after learning the rules of the game. In the end, Google’s self-taught artificial intelligence beat the vaunted Stockfish 8 with 28 wins, 72 draws, and no losses whatsoever, in a performance DeepMind labeled “superhuman.”
Chess Grandmaster Peter Heine Nielsen says that he had “always wondered how it would be if a superior species landed on earth and showed us how they played chess.”
“Now I know.”
Follow Nate Church @Get2Church on Twitter for the latest news in gaming and technology, and snarky opinions on both.