AI systems now write formal proofs and Lean4 is the software trusted to certify them — but its own creator went from celebrating its rigor to publishing postmortems for kernel soundness bugs, warning that AIs are good at exploiting them. Daniel and Emily ask what a certificate is worth when the verifier itself is young, unproven software, and what humans still have to do.
## Hosts
### Daniel Altman
Daniel Altman is a computer scientist and AI strategist with a background in engineering, quantitative research, and emerging technologies. He studies how technological breakthroughs reshape industries, markets, work, and human capabilities.
### Emily Carter
Emily Carter is a business strategist and former technology product leader with experience building digital products and studying how innovation reaches millions of users. She focuses on AI adoption, business transformation, work, education, entrepreneurship, and leadership.
## Sources
- https://www.lesswrong.com/posts/zHhGa3iQBQcdSifFL/a-critical-look-at-lean4-1
## Episode Information
Language: en
Published: 2026-10-12T20:21:05.960883+00:00
