Are We Stuck with Lean?

A mathematician questions whether the community is locked into Lean as the dominant interactive theorem prover (ITP), given that its popularity may owe more to prominent adopters (Kevin Buzzard, Peter Scholze, Terry Tao) than to objective superiority. The post raises concerns about Lean's soundness bugs and its propositions-as-types foundation, and proposes Metamath as an alternative with higher correctness assurance and a set-theoretic basis. The author also notes that AI's growing ability to write formal mathematics could make it feasible to build a Mathlib-equivalent for another ITP, though institutional support would be needed. The post stops short of advocating abandoning Lean, instead calling for a viable alternative with stronger soundness guarantees.

Source: https://mathoverflow.net/questions/513742/are-we-stuck-with-lean. 8sync News only summarizes and links out; content copyright belongs to the authors and original sources.