DeepMind’s AlphaGo: Google Breakthrough in Artificial Intelligence



AlphaGo beats human

In January 2016, Google (GOOGL) revealed its AlphaGo game-playing software, which had already beaten the professional European Go Player Fan Hui in October 2015. AlphaGo is an AI (artificial intelligence) software product developed by Google’s subsidiary DeepMind. AlphaGo learns by itself based on the experiences gained over the period.

Google is running several versions of this software, which helps AlphaGo play unlimited games against itself. As a result, it allows the software to experience and acquire knowledge at a faster rate than a human can.

Article continues below advertisement

AlphaGo leaves Deep Blue behind

Recently, Google’s (GOOG) DeepMind has made a breakthrough in artificial intelligence by beating Lee Sedol, the world’s best Korean player of Go. DeepMind’s success came as a surprise for experts who believed that AlphaGo would take around ten years to beat the top-ranked professionals players of Go. The search engine giant has already deployed AI in its range of products such as automatically writing e-mails, suggesting YouTube videos, and driving an autonomous car.

AlphaGo is a computer program that uses millions of connections that mimic the structure of the human brain to recognize patterns and make intuitive judgments. DeepMind also plans to devise the existing algorithm that requires less knowledge of the game by changing the Go board (a square of 19-by-19 line grids), to check AlphaGo’s consistency. AlphaGo is a contrast to the IBM’s (IBM) Deep Blue, which was mainly designed to play chess in 1997.

Microsoft (MSFT) and Google (GOOG) constitute 7.4% and 4.5% of the PowerShares QQQ ETF (QQQ)


More From Market Realist