← Home
This Week in Startups · September 15, 2025 · 60m

Lightmatter: Photonic Computing for AI

Nick Harris of Lightmatter explains photonic computing — using light instead of electrons for AI computation — and why it could break the energy and heat barriers of traditional chips.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Canon

Harris explains that the physical environment of electronic computing (heat dissipation, power consumption) is becoming the primary constraint on AI progress, not algorithm design.
Every improvement in compute efficiency is immediately consumed by training larger models, meaning the demand for compute never feels satisfied regardless of hardware advances.
Was this useful?