How a functional programming language is turning the world of formal mathematics inside out
In short: Lean is a proof assistant that allows computers to verify complex mathematical theorems with absolute certainty. By digitizing research-level mathematics and physics, it has moved from an experimental tool to a powerful engine for AI-assisted discovery, recently enabling the formalization of Fermat's Last Theorem in just eleven days.
Launched in 2013 by Leonardo de Moura, Lean is built on the Calculus of Inductive Constructions, a foundational type theory. While early versions were experimental, the 2021 release of Lean 4 marked a significant shift. By allowing the system to produce C code and implementing its own frontend in Lean, the software gained the efficiency and flexibility needed for domain-specific automation. Today, it is supported by the Lean Focused Research Organization, which focuses on scaling the language for broader use in mathematics and verification.
The power of Lean lies in its ability to bridge human intuition with machine-checked rigor. The community-maintained library, mathlib, has formalized over 210,000 theorems and 100,000 definitions as of May 2025. This repository serves as a massive, cohesive foundation for pure mathematics, attracting high-profile mathematicians like Terence Tao and Kevin Buzzard. Beyond pure math, the ecosystem is expanding into physics through projects like PhysLib and software verification, where AI models are now being used to generate and check proofs, sometimes even uncovering errors in published scientific papers.
The integration of AI has accelerated Lean's capabilities dramatically. In 2024, Google DeepMind’s AlphaProof achieved silver-medal performance on International Mathematical Olympiad problems, and by 2026, AI agents had produced a machine-checked formalization of Fermat's Last Theorem. While these milestones demonstrate the potential for machines to handle massive formalization tasks, the field remains cautious. As seen when a purported disproof of the Collatz conjecture was invalidated after a bug fix, the integrity of the system depends entirely on the correctness of the underlying code and logic.
Ultimately, Lean is changing how we define mathematical truth. By providing a platform where every step of a proof is verified by a computer, it eliminates ambiguity. Whether through interactive tutorials like The Natural Number Game or large-scale collaborative projects, Lean is positioning itself as the standard for formalizing human knowledge, pushing toward a future where machines not only verify our work but help us navigate the frontiers of theoretical physics and mathematics.
Opens the Intellect+ feed with this card first. Swipe up for the next one.
Sources
- Summary by Intellect · source: Lean (proof assistant)
To cite this card, link to https://intellectplus.co.uk/app/technology-lean-proof-assistant.
How we source, check and correct cards: Editorial standards.
General information, not professional advice. Spotted an error? Flag the card or see corrections and complaints.
Intellect Plus is a visual educational discovery platform for exploring fascinating facts and ideas. Brain Food, Not Brain Rot. What is Intellect Plus?
Related
More in Technology today · More in Technology · Latest · A–Z · All topics