Super excited to introduce SayCan (say-can.github.io): 1st publication of a large effort we've been working on for 1+ years
Robots ground large language models in reality by acting as their eyes and hands while LLMs help robots execute long, abstract language instructions
