formal methods

Evidence (6)

• 10Hacker News·9/4/2026breakthrough
  Claude Formalizes Fermat’s Last Theorem

  Anthropic says Claude autonomously produced the first end-to-end, computer-checked formal proof of Fermat’s Last Theorem in 11 days. The proof was written in Lean and involved millions of lines of code and tens of thousands of intermediate theorems, highlighting progress in AI-assisted formal verification for mathematics.

• 7The New Stack·8/19/2026research
  Canonical tests AI-assisted C-to-Rust translation on security-critical code

  Canonical is funding researchers at the University of Bristol to evaluate whether automated tools can translate large C codebases into Rust without changing behavior. The project focuses on security-sensitive components such as AppArmor and snap-confine, and combines language-model code generation with fuzzing, formal analysis, and symbolic repair to verify equivalence and reduce unsafe usage.

• 4Hacker News·7/30/2026research
  Why formal methods remain underused in software

  This article gives a historical and conceptual overview of formal methods, explaining why they have seen limited adoption despite their promise for correctness. It distinguishes between specification and verification, and between code and design, while arguing that the main barriers are practical complexity, cost, and fragmented terminology rather than a simple lack of usefulness.

• 5Martin Fowler·7/13/2026research
  Fowler on AI Harnesses and Self-Hosted Models

  Martin Fowler reports that teams are focusing more on AI harness engineering, especially context management and stronger validation techniques such as property-based testing and formal methods. He also notes growing interest in self-hosted open-weight models, driven by cost, data security, model sovereignty, and the possibility of fine-tuning domain-specific models to reduce token usage.

• 4InfoQ·7/10/2026research
  Podcast on Formal Methods in an AI-Powered Future

  In this InfoQ podcast, Shane Hastie speaks with Gabriela Moreira about making formal methods more accessible through the Quint specification language. The discussion also covers how AI is lowering the barrier to formal specification and model-based testing, while emphasizing that defining correct system behavior remains a human responsibility.

• 4Hacker News·6/29/2026research
  Questioning what software can truly guarantee

  This ACM Queue article argues that software should be evaluated by the guarantees it can actually provide, rather than by broad assumptions or marketing claims. It discusses how correctness, reliability, and safety depend on explicit specifications, testing, and system boundaries.