RoboBoom
NEXT with John Koetsier· August 26, 2026· 18:37

Made in America Robots: 3X more expensive than China?

Jan Liphardt, founder and CEO of Open Mind and Stanford bioengineering professor, discusses U.S. readiness to build physical AI with NEXT host John Koetsier.

In this talk

  1. 0:09

    The cost and supply challenge

    Open Mind’s Physical AI Readiness Index estimates that building a humanoid without Chinese suppliers could cost 2.85 times as much, turning a $50,000 robot into one costing about $150,000. Liphardt says the challenge also includes securing materials and components such as rare earths, magnets, gears, actuators, and chips.

  2. 2:34

    Rebuilding a scattered industrial base

    Liphardt says U.S. manufacturing capacity is spread across the country, with relatively few suppliers for some components. He contrasts this with Shenzhen, where businesses making parts such as bearings, sensors, and circuit boards are clustered nearby, enabling faster design cycles. He suggests existing automotive capabilities could support advanced robotics.

  3. 7:09

    Hardware shapes the global rankings

    Liphardt ranks China first, Japan second, and the U.S. third in physical AI readiness, citing Japan’s strength in actuator engineering. He argues that access to hardware supply chains and the ability to integrate components matter alongside AI models. China leads the index with 83 points, compared with 69.6 for Japan and 69 for the U.S.

  4. 11:33

    Fast feedback can speed development

    Liphardt points to Ukraine’s drone production as an example of rapid iteration, with design changes informed by frontline experience in days or weeks. He says U.S. states could similarly support real-world robot trials in settings such as hospitals or infrastructure projects, giving engineers useful feedback and training data.

  5. 14:26

    Integration and data are strategic advantages

    Liphardt describes the U.S. lead in AI as narrow and says China’s large manufacturing base could give it far more physical AI training data. He sees vertical integration, as pursued by Tesla and Optimus, as strategically sensible despite the difficulty of making cost-effective robots that way. Koetsier also discusses prices ranging from 1X’s announced $20,000 price to six-figure systems.

Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.

More talks