Google’s been making a lot of headlines in recent days with regards to artificial intelligence (AI). The company’s UK-based DeepMind division (an AI company that Google purchased in 2014) developed an AI-powered computer program named AlphaGo, whose sole purpose is to play and master the ancient Chinese board game of Go. The program uses a set of highly advanced algorithms in order to determine the best possible move versus an opponent, and while it’s not the first AI-program designed for such a purpose, it’s the first to actually beat a world champion Go player without any handicaps.
Black Friday 2017 Deals: Find Great Deals on Android Smartphones, TV’s, Smart Speakers, Chromebooks and More.
AlphaGo made history when it beat one of the world’s top ranked Go champions, Lee Sedol (currently ranked second internationally in Go championships) in the first of five planned matches. The AI program went on to win the second and third matches as well, leading many to wonder if it could pull off an historic sweep. Sedol put that speculation to rest, however, when he finally beat AlphaGo in the fourth match of the series. While the South Korean Go champion was understandably relieved to have avoided a complete defeat at the hand’s of Google’s AI phenomenon, AlphaGo still takes the series crown regardless of the final match’s outcome. The fact that an AI program was able to defeat one of the world’s greatest Go champions three times in a row is still a remarkable feat, and it highlights DeepMind’s remarkable progress in developing AI technology.
For those unfamiliar with the finer points of Go, the rules are actually very simple, but the number of possible outcomes is staggering. Without going into an overwhelming amount of detail, the game involves two players who have a collection of stones, and who then take turns placing those stones on a game board. Winning the game essentially boils down to whichever player can cover the majority of the space on a game board’s total area. As mentioned earlier, the number of possible permutations on a final game board is overwhelmingly high, and companies have struggled for years to develop AI programs that could challenge the very best Go players in the world. AlphaGo’s accomplishment is an historic milestone for the AI field, showing the entire world just how much the technology has matured in recent years.