Google’s Al Computer Beat the Best Go Player

go5Early this year the final game of the historic match between Google’s artificially intelligent Go-playing computer and Korean Go Grandmaster Lee Sedol was won by Google’s AI computer, taking the best-of-five series with one loss and four wins.

The win placed an exclamation mark on a noteworthy occasion for artificial intelligence. In the last twenty-five years, computerized machines have defeated the best humans at othello, chess, checkers, and jeopardy. However, this was the first time a device surpassed the very finest at Go, a 2,500-year-old game that is more intricate than chess, and that requires an enhanced degree of perception of the human mind.

A Crucial Leap Forward

The win was hailed as a crucial leap forward for machine intelligence. Because it is a very complex game, Go frequently depends on intuition. The victory of an AI computer showed that computers are well on their way to attaining the mental powers that we humans imagined belonged only to us.

AI professionals had initially thought that it would require another decade of progress for computers to get good enough to defeat a human player. However, the AlphaGo system beat the European champion the previous year, and eventually managed to beat Lee, the world’s greatest Go player. Consequently, experts have alleged that the computer’s victory has evolved AI technology by at least a decade.

DeepMind

go6Lee Sedol is an exceptional player, and he was too strong for AlphaGo throughout the competition. After he had won one game over Alpha Go, the DeepMind software team deliberated on and analyzed the loss when they returned to the United Kingdom, to figure out which issue caused the computer to lose.

AlphaGo played a remarkable game, but Sedol played a spectacular move on turn 78 that took the computer by surprise. All through the competition DeepMind’s computer made unusual moves that a human player characteristically wouldn’t contemplate. Some observers even questioned whether they were errors, but they were just very rare moves.

However, it was said that the AlphaGo program made one strange move on turn 79 of the game. The AlphaGo co-founder Hassabis tweeted saying that the move was a factual inaccuracy, rather than an unusual move. He also likened the success in the first game to landing on the moon, since 60 million people observed the moment in China alone.

Perpetual Fears

In spite of the innumerable superlatives used to define the outcome of the first two games, there have also been those who have not applauded the breakthrough. The perpetual fears about the progression of artificial intelligence led some, although in a rather light note, to pronounce Lee’s defeat as a bad day for the humanity.