CES® 2027 NEWS
ROAD TO CES 2027Live now
ROAD TO CES · AN INTELITOYS EVENTJAN 6–9 · VEGAS
Home Robots

Tutor Intelligence Says Human Tutors, Not Lab Breakthroughs, Will Scale AI Robots

CEO Josh Gruenstein argues the robot intelligence bottleneck can be broken by connecting deployed robots to remote human tutors who train foundation models from real work.. Tutor Intelligence says its platform deploys robots ahead of full autonomy and uses remote human tutors to train robot foundation models from real warehouse and factory work, an approach the company…

InteliToys Editors ? Tutor Intelligence Tutor robotics platform (Cassie and Sonny)
Tutor Intelligence co-founders with the Sonny and Cassie robots in a warehouse (image: Tutor Intelligence)

Tutor Intelligence has published a founding essay by co-founder and CEO Josh Gruenstein laying out the company's case for scaling AI robotics by connecting deployed robots to remote human tutors rather than waiting for a lab breakthrough in autonomy. The company says the robotics field has coalesced around the belief that intelligence is the bottleneck to widespread robot proliferation, and that its answer is to deploy robots now and build intelligence one action at a time.

Tutor describes itself as a robotics research and deployment company that builds generally capable robot hardware, deploys ahead of autonomy, and uses human tutors to train robot foundation models from real work. The company says owning every layer - robots, models and deployments - allows the platform to compound as a network, with every robot put to work making the next robot smarter, cheaper and easier to deploy.

Gruenstein writes that tutors let robots create positive net economic value immediately, because breaking the intelligence bottleneck allows robots to be placed into tasks and industries that do not pencil out with specialized industrial automation. He argues that anyone deploying a large volume of tutored robots would sit in the path of a massive flow of robot demonstrations, creating what he calls a new internet of robot training data that could train internet-scale robotics models.

Ride the Las Vegas Monorail to CES - stops at the Las Vegas Convention Center

The company says its first robot is Cassie, and that heavy, bulky materials like boxes and pallets are the protocol buffers connecting the physical world, moved today almost entirely by human workers. Tutor says Cassie runs manufacturing and logistics workflows for customers ranging from small family-owned factories to Fortune 50 distributors across the United States.

Tutor says it was founded by Gruenstein and Alon, and that the company's model treats hardware, intelligence and customers as a co-design problem. The company argues that profits from solving a customer problem can be reinvested in each layer, whether training data for a foundation model or manufacturing capacity for robot hardware, unlocking more data and supply chain scale over time.

The essay is published on Tutor Intelligence's blog alongside the company's announcement of a $75 million Series B funding round led by a16z.

Source: Tutor Intelligence.

InteliToys is editorial. We do not sell Tutor Intelligence Tutor robotics platform (Cassie and Sonny). Watch the source tape in the branded lightbox below.