Content-type: text/html Downes.ca ~ Stephen's Web ~ Reinforcement learning challenge to push boundaries of embodied AI

Stephen Downes

Knowledge, Learning, Community

This article (and also this one, in the fourth point) discusses the idea that intelligence is more than just brain states. This is known generally as the concept of 'embodied cognition', and hence in AI, as 'embodied AI'. "Human and animal brains have evolved along with all other body organs with the ultimate goal of improving chances of survival. Our intelligence is tightly linked to the limits and capabilities of our bodies." In this article, the concept of embodies AI is put to the test using reinforcement learning for navigation tasks in 3D worlds. It seems simple; we do it all the time. In fact, though, it's the result of a complex interaction of human vision and agency. This helps explain why practice and hands-on activities are so important for human learning.

Today: 9 Total: 1120 [Direct link] [Share]


Stephen Downes Stephen Downes, Casselman, Canada
stephen@downes.ca

Copyright 2024
Last Updated: Apr 25, 2024 11:17 p.m.

Canadian Flag Creative Commons License.

Force:yes