All Recordings

The Physical Turing Test: Solving General Purpose Robotics

About This Talk

Explore what it takes to evaluate and benchmark general-purpose robotic systems. NVIDIA's Jim Fan discusses Project GR00T, embodied AI agents, and how humanoid robotics requires new evaluation methods beyond traditional benchmarks. Featured at Actuate, Foxglove's annual conference for robotics developers.

Speaker

Jim Fan, Distinguished Scientist, NVIDIA

Dr. Jim Fan is the co-Lead of Generalist Embodied Agent Group (GEAR) at NVIDIA. GEAR’s mission is to build generally capable Physical AI. Dr. Fan’s team is spearheading Project GR00T, NVIDIA’s moonshot initiative at building the AI brain for humanoid robotics. Previously, Dr. Fan obtained his Ph.D. degree at Stanford Vision Lab. His publications have won several awards and nominations, including the Outstanding Paper Award at NeurIPS 2022. He spearheaded multiple impactful works on robot foundation models, AI gamers that play Minecraft proficiently, and general-purpose agents that learn to operate a computer. His works have been widely featured in news media, such as New York Times, Forbes, MIT Technology Review, TechCrunch, The WIRED, VentureBeat, etc.

More Recordings