This podcast interviews Vincent Vanhoucke, a distinguished engineer at Waymo, about the impact of LLMs on autonomous vehicles and robotics. The conversation covers how LLMs enhance Waymo's self-driving technology by providing "world knowledge" and improving model training, while also discussing limitations and the need for safety verification layers. Vanhoucke also shares insights into his transition from Google's robotics team to Waymo, highlighting the similarities and differences between the two fields. Key challenges in autonomous driving, such as addressing the "long tail" of rare events through simulation, and the future role of sensors are explored. The interview concludes with reflections on the state of robotics and the potential of LLMs to revolutionize the field.
