Published on June 3, 2018 by

A recent paper called ‘world models’ has gotten really popular in the machine learning community. They trained an AI to play a racing game by having it learn inside of its own simulated dream environment. Meaning, the AI learned a model of what the game world was like, then was able to generate a game world that was roughly similar to what it learned and train inside of that. A simulation inside of a simulation. I’ll explain how their model was structured both theoretically and programmatically in this video.

Code for this video:
https://github.com/llSourcell/world_models

Please Subscribe! And like. And comment. That’s what keeps me going.

Want more education? Connect with me here:
Twitter: https://twitter.com/sirajraval
Facebook: https://www.facebook.com/sirajology
instagram: https://www.instagram.com/sirajraval

More learning resources:
https://worldmodels.github.io
https://arxiv.org/abs/1803.10122

Join us in the Wizards Slack channel:
http://wizards.herokuapp.com

Sign up for the next course at The School of AI:
https://www.theschool.ai

And please support me on Patreon:
https://www.patreon.com/user?u=3191693

Category Tag