News & Reviews:   google deepmind

If you've grown up playing StarCraft, you'd remember how terrible the A.I. is back in the 90's, with their units hilariously walking in a straight line and not knowing what to do after the first two waves of attack; or retreating their entire army back to their base in StarCraft II if one of their workers is harassed, despite the improved A.I.. Well, that could be all changed in the future, now that Google Deepmind has announced their collaboration with Blizzard Entertainment to incorporate A.I. and Machine Learning in StarCraft II with researchers around the world.

Yesterday, we reported that a Go-playing computer program, AlphaGo, grabbed the first win against the best Go player in the world, Lee Sedol and it was the first time that a system could do such feat in human history. Today, it got the second win again this time after 211 moves. Both players played two full hours of regulation time, going into byō-yomi overtime. And just for your information, AlphaGo actually learns from experience. 

If you've never heard of Google DeepMind, then welcome to the right place to get yourself educated. DeepMind is an artifical intelligence designed by Demis Hassabis, Shane Legg and Mustafa Suleyman in London during the year 2010, its purpose it to solve intelligence and use it to make the world a better place and was acquired by Google in 2014.

And just yesterday, during its first game at the Google DeepMind Challenge match, the system has defeated the world's best Go player, Lee Sedol after 186 moves. This is the first time in history that a computer program has bested a top-ranked human Go player on a full 19x19 board with no handicap.