Episode 36 · Engineering Deep Dives
Type Theory and Formal Verification: The Clerk Who Never Gets Tired
Testing a program is different from proving a property about it. This episode starts with types as contracts and the value of making units explicit, then develops the correspondence between proofs and programs. Trusted kernels and proof assistants illustrate how complex arguments are checked against small foundations. Equality, higher structure and AI-assisted proving extend the discussion, while a return to the plant asks where formal methods can make a concrete difference.
Listen to the show on your preferred podcast platform.
Listen on Spotify ↗Listen on Apple Podcasts ↗Listen on Pocket Casts ↗Listen on Podcast Addict ↗Listen on Castbox ↗Listen on TuneIn ↗
See it work
Explore the working principle
Inside this episode
- Cold open: the trip that almost didn't happen
- Why testing can't save you
- Types as contracts, and why units matter
- Curry Howard: proofs are programs
- The trusted kernel and the de Bruijn criterion
- Rocq, Lean, Isabelle: a field guide
- CompCert: the compiler that random testing couldn't break
- seL4: proving an operating system kernel
- Proofs in production: from cloud permissions to protocols
- Four colours, odd orders, and oranges: the mathematicians arrive
- The Liquid Tensor Experiment and the three-week sprint
- Equality, univalence, and the shape of sameness
- AI meets the proof checker
- Back at the plant: what Zara actually changed
- Epilogue: the clerk who never gets tired
Production credits & AI disclosure
Scripts generated with Claude Opus. Synthetic narration generated with Edge TTS. Reviewed, listened to, and approved by Sankalp Thakur.