Skip to main content
robotics diffusion policy teaching image
TRI is developing a new method to teach robots overnight 1 Minute Read

Brian Heater from TechCrunch interviewed Toyota Research Institute’s CEO Gill Pratt on the recently announced breakthrough generative AI approach based on Diffusion Policy, a method to quickly and confidently teach robots new, dexterous skills.

“It’s remarkable how fast it works,” Pratt is quoted saying in Heater’s article. “In machine learning, up until quite recently there was a tradeoff, where it works, but you need millions of training cases. When you’re doing physical things, you don’t have time for that many, and the machine will break down before you get to 10,000. Now it seems that we need dozens.”

Read Brian Heater’s article to learn more about this new training method.