TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Google's DeepMind AI just taught itself to walk

27 pointsby Huhtyalmost 8 years ago

5 comments

anotheryoualmost 8 years ago
Clearly needs some energy constraint factored in to look less like monty python :) I also wonder if the gravity is realistic.<p>A similar video from 2013: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=pgaEE27nsQw" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=pgaEE27nsQw</a> Looks much smoother, because of muscle and nerve simulations. At 0:55 you can see the generations learning how to walk.
bkoalmost 8 years ago
If anyone is interested in learning more how reinforcement learning works, I would highly recommend openai gym [0]. It provides a controlled environment where you can develop learning algorithms to accomplish an objective goal. It also has a walkers similar to the one in the video [1]. For learning more about reinforcement learning, I would recommend Richard S. Sutton book Reinforcement Learning: An Introduction. He has a new draft available free online [2]. Or the course Practical RL [3].<p>[0] <a href="https:&#x2F;&#x2F;gym.openai.com&#x2F;envs" rel="nofollow">https:&#x2F;&#x2F;gym.openai.com&#x2F;envs</a><p>[1] <a href="https:&#x2F;&#x2F;gym.openai.com&#x2F;envs#mujoco" rel="nofollow">https:&#x2F;&#x2F;gym.openai.com&#x2F;envs#mujoco</a><p>[2] <a href="http:&#x2F;&#x2F;ufal.mff.cuni.cz&#x2F;~straka&#x2F;courses&#x2F;npfl114&#x2F;2016&#x2F;sutton-bookdraft2016sep.pdf" rel="nofollow">http:&#x2F;&#x2F;ufal.mff.cuni.cz&#x2F;~straka&#x2F;courses&#x2F;npfl114&#x2F;2016&#x2F;sutton-...</a><p>[3] <a href="https:&#x2F;&#x2F;github.com&#x2F;yandexdataschool&#x2F;Practical_RL" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;yandexdataschool&#x2F;Practical_RL</a>
honestoHeminwayalmost 8 years ago
We are pretty good at walking and moving fast forward..<p>I wonder how fast a AI could get at climbing- walking alterations, when given a challenging terrain..
peponalmost 8 years ago
one more sensationalist headline for deepmind<p><a href="https:&#x2F;&#x2F;www.google.es&#x2F;search?q=mujoco+learning+to+walk#q=mujoco+learning+to+walk&amp;tbm=vid" rel="nofollow">https:&#x2F;&#x2F;www.google.es&#x2F;search?q=mujoco+learning+to+walk#q=muj...</a>
lunleloalmost 8 years ago
Thank you for the vid