Verified but Not Understood: The Riemann Hypothesis and the New Epistemology of Proof
This opinion piece reflects on the epistemological implications of AI-assisted formal proof through the case of a November 2026 Lean formalization associated with Levent Alpöge’s proposed proof of the Riemann hypothesis and generalized Riemann hypothesis. Drawing a parallel with non-reductive explanation in the philosophy of chemistry, the author argues that human–AI co-invention can produce formally verified derivations whose conceptual architecture remains initially opaque, making human explanation and conceptual reconstruction more vital than ever.
Authors
- Nadji Belkheiri (ORCID: https://orcid.org/0000-0002-5499-6216)
Institutions
- Ziane Achour University of Djelfa (DZ)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-22
- DOI
- https://doi.org/10.5281/zenodo.24430324
- Primary Topic
- History and Theory of Mathematics
- Type
- preprint