Nov 2, 2026 · 46 min
Tutor bets deployment will teach robots to work
The Robots Already Working in American Factories | Tutor Intelligence
The episode examines whether putting robots into ordinary factories can solve the data, economics, and manufacturing problems blocking wider automation.
- 1Tutor treats robotics as a full-stack operating problem spanning models, hardware, manufacturing, deployment, and customer support.
- 2Human teleoperation turns difficult physical work into training data while helping robots deliver value before autonomy is complete.
- 3General-purpose robots could expand production and reshape industrial geography, but adoption still depends on practical economics and supply chains.
Don't miss
Josh Gruenstein explains how human tutors can handle a robot’s difficult physical tasks while generating the training traces needed for greater autonomy.
The brief
Josh Gruenstein and Raghu Raghuram frame useful robotics as a physical-economy problem: success requires coordinating AI, hardware, manufacturing, operations, and customer deployment.
Tutor puts robots into real factories and warehouses, using human tutors and teleoperation when autonomy fails; those interventions become data for improving the system.
The company’s strategy challenges the idea that robots must be fully autonomous before they are useful, especially for mid-market businesses without large automation teams.
Tutor is building its own robot platforms because existing industrial supply chains are too narrow for the ordinary American facilities where much repetitive work still happens.
The broader wager is economic: flexible robotic labor could lower costs and expand production, while domestic manufacturing and U.S.-China supply-chain ties complicate the transition.
Weekly email
State of the Pod, every Monday
The week's most-discussed names and stories from podcasts, with the moments behind them. Free, five minutes, unsubscribe in one click.