Physical AI Isn't About Bigger Models [Part 2] | Ted Lutkus Onyx Robotics

This is Part 2 of my conversation with Ted Lutkus, the Founder and CEO of Onyx Robotics. Robotics and physical AI get challenging in the real world because we start dealing with unmodelled forces. This is a non-trivial and expensive problem across anything that moves: cars, drones, collaborative arms, humanoids and much more. Onyx is developing the controls and simulation layer and the tooling to empower controls engineers. We talk about how to train your controller, foundation models, data and real world testing.Whatever your background, I’m sure you’ll have a lot to take away!00:00 - Intro01:59 - Train With Onyx: Product Walkthrough06:21 - On Foundation Models8:38 - Real World Testing 10:03 - Why Build In RoboticsCheck out Onyx Robotics: https://www.onyx-robotics.com/Follow Ted Lutkus: https://www.linkedin.com/in/ted-lutkus-787904137/______ References:Marco Hutter & ANYMal, ETH Zurich: https://arxiv.org/abs/2505.16084 https://raisim.com/sections/ArticulatedSystem.htmlhttps://www.youtube.com/watch?v=fJfAWiylpxwSystem Identification: https://www.mathworks.com/help/ident/gs/about-system-identification.html