Sergey Levine discusses Physical Intelligence robot foundation models
Aditya Agarwal shared Sergey Levine's conversation at SPC: In his early years, Levine had a roomful of robots in a converted Google conference room spend all day grasping objects, accumulating a great deal of experience yet lacking prior knowledge; this bottleneck ultimately gave rise to Physical Intelligence. The conversation covered the Pi Zero robot foundation model, the evolution of reasoning and generalization capabilities from Pi 0.5 to Pi 0.7, as well as the robot data challenge and world models.
Body
“The universe is really big, and almost none of it is inside computers.”
When a problem is hard, make it bigger. @svlevine put a room full of robots in a repurposed Google conference room and had them grasp objects all day.
It worked. Then it hit a wall: lots of experience, no prior knowledge.
That wall became @physical_int.
This conversation at @spc was awesome. I learned a lot!
(00:28) From Video Games to Robotics: Sergey Levine's Journey (04:24) Founding Physical Intelligence: Why the Timing and Team Mattered (06:30) Inside Pi Zero: Building a Foundation Model for Robots (09:37) From Pi 0.5 to Pi 0.7: Teaching Robots to Reason and Generalize (16:45) What Makes a Great Robotics Researcher? (21:10) Solving Robotics' Data Problem and Building Learning Feedback Loops (28:04) The Future of Robotics: General-Purpose AI vs. Specialized Robots (33:24) World Models, Data Strategy, and Building a Robotics Research Culture