Episode Details

Step inside Figure’s campus to see how humanoid robots move from lab prototypes to factory-ready workers. (Original episode ~45 minutes; condensed to ~12 minutes.) Host Molly O’Shea walks the floor with founder Brett Adcock, revealing Helix—the onboard vision-language-action model—alongside the manufacturing spaces BotQ, The Grid, and their Never Fall safety program. Learn how Figure combines neural-first control, reinforcement learning, sim-to-real transfer, custom batteries-as-structure, and rigorous validation to produce persistent, local autonomy for homes, logistics, and factories. Key takeaways: why onboard inference matters, how robots learn resilience to hardware failures, what large-scale data and anonymization mean for deployment, and how multidisciplinary manufacturing enables scale. If you want a concise, technical look at humanoid robots, AI autonomy, reinforcement learning, and real-world robotics manufacturing, this summary delivers the essentials. Listen now to get the key ideas in minutes.