On: Reinforcement learning with musculoskeletal models

[tedivm]

The skeleton is not exactly good marketing.

[stcredzero → tedivm]

I was going to suggest that they hook the AI up to QWOP.

[tlrobinson → stcredzero]

I was curious if anyone had tried to train an AI to play QWOP. Of course they have: https://www.youtube.com/watch?v=e27TUmMkOA0 (among others)

I wonder if you could get a better result by including other factors in the reward function, like trying to maintain a slight forward lean.

[mxwsn → stcredzero]

But this task is strictly harder and more general than QWOP .

[juancb → mxwsn]

But not strictly as humourous.

[stcredzero → mxwsn]

For PR purposes.