Skynet is here.
Google's AI project Alphazero taught itself to play chess, and beat the world's best chess player 24 hours later. 😱
Typically, computers have been taught to play games like chess and Go via a combination of human expertise and computer speed. According to a new paper, the latest incarnation of AI was only taught the rules of chess and left to play against itself.
Within 24 hours, the computer had achieved superhuman capabilities at chess.
Interestingly, the computer also developed a lot of chess opening theory that humans have developed over hundreds of years.
The computer appeared to prefer 1.d4 and 1.c4, loves the Queen's Gambit, rejected the French Defense, never liked the King's Indian, grew disillusioned with the Ruy Lopez, and surprisingly never fell in love with the Sicilian Defense.