New interview series dives into Lean formal methods with Logical Intelligence's Tanner Duve

A newly launched interview series for the Lean and formal methods community has published its first episode, featuring Tanner Duve, a Member of Technical Staff at Logical Intelligence. Duve works on formal verification and compilers in Lean, the functional programming language and proof assistant, and is an open-source contributor to Mathlib and CSLib. The conversation spans 1 hour and 20 minutes, covering everything from the social dynamics of formal verification to the technical weeds of free monads, Turing (in)completeness in Lean, and AI-assisted math formalization. Notably, Duve also shares his unusual background as a former Division 1 football player, a detail that adds a human angle to an otherwise deeply technical discussion.

The interview is structured with clearly marked chapters, making it easy to jump to specific topics—including AlgoLean, his thoughts on Lean versus Rocq and Haskell, Rust's type theory, and the pedagogy of programming language theory. Duve also reflects on how formal verification has changed pre-AI versus now, a timely perspective given the rapid growth of AI tools in proof assistants. For developers and researchers in formal methods, the episode offers both strategic insight and tactical advice: how to contribute to Mathlib, where to learn type theory, and how to think about the future of AI in math. Whether you're a Lean veteran or just curious about where formal verification is heading, this interview is a compelling starting point for a promising series.