Useful robots won’t be programmed one task at a time; they’ll need to adapt to unfamiliar objects, environments, and robot bodies. Physical Intelligence cofounder Chelsea Finn (named last month to the TIME100 AI list for 2026) joins Reid Hoffman and Aria Finger to explain how general-purpose robot models learn to act in the messy physical world. She explores why lower-quality training data can strengthen a model, what months of failed laundry-folding attempts revealed, and how the same system can work across different robot platforms. A robot mistaking an oven for a drawer becomes a lesson in the strange line between useful generalization and obvious error. They also examine why robot demos can mislead, how hardware reliability can stall progress, and what it will take to move AI off the screen and into the world.
Congratulations to Chelsea, recently named one of TIME’s 100 most influential people in AI:
Podden och tillhörande omslagsbild på den här sidan tillhör
Reid Hoffman. Innehållet i podden är skapat av Reid Hoffman och inte av,
eller tillsammans med, Poddtoppen.
