…2015 that the first AI Go player was able to beat one of the best human players in the world. Then, just last year, AlphaGo beat the top human player in the world in three out of three matches. It had taken years of development, feeding millions of…

…It had taken years of development, feeding millions of moves by hundreds of thousands of human players into its database, for AlphaGo to be able to defeat the best human player in the world. Then something new happened. Google's DeepMind AI was set…

…best human player in the world. Then something new happened. Google's DeepMind AI was set the task of creating a new version of AlphaGo. But without using any human input, it created this new version using only the rules of go and playing itself. In just…