Quasi-Riemann hypothesis proved by AI - mathematician explains | Lex Fridman
A mathematician discusses the Riemann hypothesis and recent breakthroughs in AI-assisted mathematical proofs, including the quasi-Riemann hypothesis proof and the formalization of Fermat's Last Theorem in Lean, which exceeded his expectations for what AI systems could accomplish.
Summary
The transcript features a discussion between Lex Fridman and a number theorist about major developments in mathematics and AI. The mathematician explains the Riemann hypothesis by describing how Riemann discovered that a complex function (the Riemann zeta function) governs the distribution of prime numbers. He uses an analogy to signal processing and audio engineering to explain how the zeta function analyzes prime numbers similar to how a microphone analyzes sound waves. The Riemann hypothesis posits that all the "amplitudes" (zeros of the zeta function) have equal value, like an orchestra where every instrument plays at the same amplitude.
The conversation then shifts to recent AI breakthroughs that shocked the mathematician. He recounts being at a National Museum of Mathematics gala when the quasi-Riemann hypothesis proof was announced, and his immediate disbelief at seeing it in the AI system's output. He explains his previous understanding of AI capabilities in mathematical formalization, stating that AI could only build one or two layers on top of existing Mathlib (a library of formally verified mathematics) through "jerry-rigging," but struggled with "canonization"—refactoring code to match library standards.
The mathematician then discusses the formalization of Fermat's Last Theorem in Lean, which required 13 million lines of code and completely contradicted his expectations from a colloquium he had given just two days prior. He describes how Anthropic's AI system blasted through all the difficulties, producing a sorry-free (complete) proof through multiple layers of jerry-rigging rather than elegant, library-quality code. He estimates only 2-3 humans were involved, with their main role being to say "Keep going" in all caps. He emphasizes that while the proof isn't clean or beautiful, it demonstrates that autoformalization is essentially solved as a problem. The mathematician concludes by noting this was the third major mathematical breakthrough in a month (after Fermat's Last Theorem came Navier-Stokes), expressing both excitement and a sense of being overwhelmed by the pace of progress.
Key Insights
- The Riemann hypothesis predicts that all the zeros of the zeta function have equal amplitude, which can be understood as an orchestra playing the 'music of prime numbers' where every instrument sounds at exactly the same amplitude
- AI systems were previously limited to building one or two layers on top of existing Mathlib through jerry-rigging, struggling with canonization (refactoring code to match library standards and elegance)
- The formalization of Fermat's Last Theorem in Lean with 13 million lines of code was accomplished with only 2-3 humans involved, whose primary role was providing encouragement to continue, not domain expertise or pathway creation
- The mathematician states that 'autoformalization is basically solved' as a problem in terms of getting sorry-free proofs, though the resulting proofs are not elegant or library-quality and further work is needed to make them beautiful
- Three major mathematical breakthroughs (quasi-Riemann hypothesis, Fermat's Last Theorem, and Navier-Stokes) occurred within a month, all far exceeding the mathematician's previous model of what AI systems were capable of achieving
Topics
Transcript
[0:02] but first, we'll have to break if it's okay. So Alex, you're a number theorist. Um, maybe it's fair to say the central problem in your field is the Riemann hypothesis, and you showed a lot of excitement on the aforementioned Twitter/X about the quasi-RH, the quasi-Riemann hypothesis proof. Can you explain that? Can you explain what the Riemann hypothesis is, and can you explain the nature of your excitement about it? - Sure. There's sort of a one-minute version of the Riemann hypothesis- [0:35] ... or a five-minute version or a thirty-hour version. I don't know how far you wanna, you wanna go. - Let's start pulling at that string. - So we know what the prime numbers…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Lex Clips
Playing chess in prison vs serial killer | Larry Lawton and Lex Fridman
Larry Lawton describes witnessing an exceptionally skilled chess player in county jail who appeared to be a Russian murderer. The player demonstrated world-class ability by defeating multiple opponents simultaneously in just 10-13 moves while barely looking at the boards, prompting discussion about how elite chess players can visualize entire games mentally.
GTA 6 review by biggest jewelry thief in US history | Larry Lawton and Lex Fridman
Larry Lawton, a legendary jewelry thief and FBI's most-wanted criminal, discusses his extensive experience robbing stores while reviewing GTA 6's trailer with Lex Fridman. He provides detailed critiques of the game's accuracy regarding criminal operations, casing techniques, and security vulnerabilities, while reflecting on how his past crimes cost him years with his children.
Robberies in GTA 6 - jewelry thief explains | Larry Lawton and Lex Fridman
Larry Lawton, a former jewelry thief, discusses his criminal expertise with Lex Fridman and explains what he wants to see portrayed accurately in GTA 6's robbery mechanics. He emphasizes that successful robberies depend heavily on meticulous casing and understanding human behavior patterns rather than just technology.
US war against China or Russia: Why military experts are wrong | Lex Fridman Podcast
Military analysts often make prediction errors by abstracting military power as a simple ranking rather than understanding how forces interact in specific contexts. The Ukraine-Russia war exemplified this, as analysts underestimated Ukrainian military capability and overestimated Russian performance, partly due to having more intelligence on Russian plans than Ukrainian defenses.
Is Putin being lied to? - The distortion of information in Ukraine war | Military Analysis
The transcript discusses how Russian military forces systematically overstate territorial control through false reporting, creating a cascading problem where fabricated battlefield data leads to poor operational decisions, increased casualties, and potentially misinforms political leadership including Putin. The speakers contrast this with Ukrainian practices and explore broader patterns of information distortion in military hierarchies across historical conflicts.