{"id":12142,"date":"2022-09-01T22:09:19","date_gmt":"2022-09-01T22:09:19","guid":{"rendered":"http:\/\/TheNextWeb=1391164"},"modified":"2022-09-01T22:09:19","modified_gmt":"2022-09-01T22:09:19","slug":"forget-chess-deepminds-training-its-new-ai-to-play-football","status":"publish","type":"post","link":"https:\/\/www.londonchiropracter.com\/?p=12142","title":{"rendered":"Forget chess, DeepMind\u2019s training its new AI to play football"},"content":{"rendered":"\n
<\/div>\n

Researchers from DeepMind, the UK\u2019s juggernaut AI lab, have forsaken the noble games of chess and Go for a more plebeian delight: football.<\/p>\n

The Google sister company yesterday published a research paper and accompanying blog post detailing its new neural probabilistic motor primitives (NPMP) \u2014 a method by which artificial intelligence<\/a> agents can learn to operate physical bodies.<\/p>\n

Per the blog post<\/a>:<\/p>\n

\n

An NPMP is a general-purpose motor control module that translates short-horizon motor intentions to low-level control signals, and it\u2019s trained offline or via RL by imitating motion capture (MoCap) data, recorded with trackers on humans or animals performing motions of interest.<\/p>\n<\/blockquote>\n

\n
<\/p>\n

Sign up to the TNW Conference newsletter<\/h4>\n

And be the first in line for ticket offers, event news, and more!<\/p>\n

<\/a><\/div>\n<\/div>\n

Up front:<\/strong> Essentially, the DeepMind team created an AI system that can learn how to do things inside of a physics simulator by watching videos of other agents performing those tasks.<\/p>\n

And, of course, if you\u2019ve got a giant physics engine and an endless supply of curious robots, the only rational thing to do is to teach it how to dribble and shoot:<\/p>\n

\n