Alright, I'm building the robot project I was putting off. This is so fucking cool.

Excellent work!

Anyone have good intro recommendations for VLAs?

(I work at Pi.)

We open-sourced Pi0 (referenced in this post): https://github.com/Physical-Intelligence/openpi

I love the implications of a robot that can plug in Ethernet cables.

Just need one that can plug in USB-A cables the first attempt (I average 3 attempts).

“Soon, a robot will fix the cables in the server room for me!”

New job title: Spaghetti Organizer

rtc handling 300ms+ delay and still pulling off tasks like plugging ethernet is kinda nuts. what i'm not getting is but how's it keeping the control loop stable without retraining? some sort of latent plan caching?

It uses an inpainting algorithm (adapted from image generation literature) to produce future actions that are consistent with the current trajectory. It's sort of like warm-starting from a cached plan, although the plan isn't latent, it's directly in action space. Hopefully that answers your question -- there are many more details in the blog post and paper :)