Google's DeepMind AI masters chess, beating expert software

Enlarge  We'd like to imagine Alpha Zero playing its chess within a 1980s style BBS

Enlarge We'd like to imagine Alpha Zero playing its chess within a 1980s style BBS

AlphaGo has come a long way since it became the first artificial intelligence to conquer the game of Go almost two years ago, but creator DeepMind Technologies, Google's AI company, does not want AlphaGo to remain a one-trick pony.

The DeepMind team admitted in their research paper that chess is "the most widely studied domain in the history of artificial intelligence", but they noted that previous chess-playing AIs required "a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades". The AlphaGo Zero algorithm was able to beat the world leading AI program, Elmo, at the Japanese board game with only two hours of self-training, winning a total of 90 games and losing only nine.

In addition to chess, the AI program was also able to win against a world champion program in other games like Shogi.

Chess isn't an easy game, by human standards.

Casting new light on eclipse eye damage
The scientists were able to use adaptive optics to obtain high-resolution images of the damage to her photoreceptors. That meant the sun's bright light was still visible and damaging to the eyes.

AlphaZero was developed at Google's DeepMind labs and is a more generic version of AlphaGo Zero, the AI that ousted the human champion of Go, a Chinese board game considered to be the most hard strategy game in the world. IBM spent more than 10 years perfecting Deep Blue before it successfully mastered chess. AlphaZero won 25 games in which it played with white (gaining the first-move advantage) and a further three in which it played black.

The research has not been peer reviewed yet.

AlphaZero developed its skills completely from scratch: as researchers write in a a pre-printed, unreviewed paper, the program was given no knowledge of chess except for the basic rules.

Men in Hungary play chess at a bath house. "It will no doubt revolutionize the game, but think about this could be applied outside chess".

Wells Fargo & Company MN Has $70.92 Million Holdings in SCANA Co. (SCG)
Edward Jones upgraded shares of SCANA from a "hold" rating to a "buy" rating in a research report on Monday, November 13th. Williams Capital lowered shares of SCANA from a "hold" rating to a "sell" rating in a report on Friday, September 29th.

The current AlphaZero takes the reinforcement learning of AlphaGo Zero and uses it in a broader approach. AlphaZero's favourite openings included the English Opening, the Queen's Gambit (my personal favourite), and the Queen Pawn Game. After eight-hours of self-play, it amassed a record of 60 wins and 40 losses against the digital old-timer.

"I always wondered how it would be if a superior species landed on Earth and showed us how they played chess", grandmaster Peter Heine Nielsen told the BBC. "I feel now I know".

Perhaps obviously, AlphaZero's dominance in chess is less impressive than its mastery over Go - a game that's significantly more complex.

This is an important point because AlphaGo has been criticised for being too narrow. But by adapting the system to learn a new set of rules for an entirely new game, the DeepMind developers demonstrated the flexibility of the system and (possibly) its potential to work outside of mere gameplay.

Gama anuncia a contratação do pentacampeão Lúcio
Com passagens de sucesso por Bayer Leverkusen, Bayern de Munique e Inter de Milão, este será o 10º clube da carreira de Lúcio . Em nota divulgada pela diretoria, o presidente, Weber Magalhães, falou sobre a chegada do atleta ao clube.

Últimas notícias