now I’m down the rabbit 🕳️ …again. currently reading https://language-to-reward.github.io 😂 thanks for sharing July

creating and destroying; flying cars, roc.camera, july.rocks

MuJoCo seems interesting - specifically MuJoCo MPC they talk about:

"combining this with a real-time optimizer, MuJoCo MPC, empowers an interactive behavior creation experience where users can immediately observe the results and provide feedback to the system..."

MuJoCo seems interesting - specifically MuJoCo MPC they talk about:

"combining this with a real-time optimizer, MuJoCo MPC, empowers an interactive behavior creation experience where users can immediately observe the results and provide feedback to the system..."

https://mujoco.readthedocs.io/en/stable/overview.html

I think https://palm-e.github.io/ PaLM-e was the eye opening moment for me

also higher levels in autonomy (motion planning and higher) have longer loops, and can work with something like LLMs to generate commands. obviously lower level controls loops are not going to be quick enough for LLMs to generate commands - which makes a lot of sense here as the author talks about