In this episode, I step onto the factory floor to examine Amazon’s robotics strategy and the growing challenge of building what experts call “physical AI.” Unlike software systems, robots must operate reliably in the unpredictable real world—handling objects, navigating spaces, and working safely alongside people.

I break down why that complexity makes robotics far harder than many AI applications, and what Amazon’s decision to shut down its Blue Jay warehouse robot project reveals about the real costs, engineering hurdles, and operational tradeoffs behind large-scale automation.