From a Worm's Brain to Mercedes: Building AI That Runs Without the Cloud // Mathias Lechner, Liquid AI

Mathias Lechner, Co-founder & CTO of Liquid AI, turned his research on the 302-neuron nervous system of the C. elegans worm into liquid neural networks and, eventually, an MIT spinoff building foundation models that run on phones, laptops and cars without the cloud. He joins Tobi to talk about why speed is non-negotiable when AI has to run locally.

He explains why Liquid AI designs its models around the hardware instead of model size, why memory bandwidth and memory size limit on-device AI as much as compute, and why he believes there's no long-term technical moat in AI, only a race where silicon partnerships give you the edge.

The conversation also covers the Mercedes in-car assistant, keeping models current in cars that live for 10 years, how the team judges training data, and his own YOLO-mode coding agent setup.