Lars Ankile

I've been thinking a lot about making robot learning scalable. To me, that means learning on the job, and using the human help still needed efficiently.

In Mulligan ⛳️, we focus each data-collection round on the initial states where the robot fails. This happens in deployment, with the robot doing the task and a person stepping in when needed.

On 3 real tasks and 2,550 blind evals, it improves on uniform collection at the same budget.

@sama the minis in the chatgpt app is an abomination, if having them at all wasn't bad enough, having them on by default is straight up insulting

and the fact that the keyboard shortcut is alt+space and cannot be turned off in the app just makes the whole thing kind of funny

finally the llm people are catching on to what we robot people have known for years

jk, really cool work from the great @wen_kaiyue and @LukeBailey181 💅

feel free to help me get my clanker cleaner faster by using my referral link: tau-robotics.com/waitlist?ref=T…

one nice things about scaling laws is that at least we know what won't work