Skip to content

The Case Against Formal Verification, 50 Years Later

Updated

Ivan Gavran revisits a 1979 critique of formal verification and asks which objections still apply. The essay traces how tools and practice have changed and returns to the role of people in deciding what a proof means for a system.

A proof checks properties stated in a model or specification. When reading a claim about verified software, ask what was proved, which assumptions the proof relies on, and how those assumptions connect to the system people use.

Ivan Gavran’s essay.

Type to search…

↑↓ navigate↵ selectEsc close