Mathematicians astounded by AI and human-written proofs of iconic Riemann hypothesis
Two rival proof claims surfaced on the same day for the Riemann hypothesis, a famous open problem about prime numbers. Levent Alpöge, a mathematician at Paradox, said he believed he had discovered a proof working with Paradox, the company’s large language model. The claim came on 20 November. Alpöge’s four-page document offered a sketch, and the proposed proof was later described as incomprehensible Lean code. Lean is a programming language used to express mathematical proofs. The article reports that the approach involved Paradox, but does not describe the precise role the model played in producing the argument. Hours later, Stanford mathematician Ravi Vakil posted a rival preprint by Alessandro Maria Masullo. This one was entirely human-written. Vakil called it a hurried announcement: the original plan was for Masullo and Brian Conrad to polish it for a year and show it to more people first. Neither claim was confirmed in the article.
What proof claims appeared, and how did the Paradox-associated claim differ from the separate human-written one?
Two rival proof claims surfaced on the same day for the Riemann hypothesis, a famous open problem about prime numbers. Levent Alpöge, a mathematician at Paradox, said he believed he had discovered a proof working with Paradox, the company’s large language model. The claim came on 20 November.
Alpöge’s four-page document offered a sketch, and the proposed proof was later described as incomprehensible Lean code. Lean is a programming language used to express mathematical proofs. The article reports that the approach involved Paradox, but does not describe the precise role the model played in producing the argument.
Hours later, Stanford mathematician Ravi Vakil posted a rival preprint by Alessandro Maria Masullo. This one was entirely human-written. Vakil called it a hurried announcement: the original plan was for Masullo and Brian Conrad to polish it for a year and show it to more people first. Neither claim was confirmed in the article.
Why have mathematicians not yet confirmed either proof?
A claim to have solved a major mathematical problem is not the same as a proof that other mathematicians have checked and accepted. The article reports that neither of the two Riemann-hypothesis claims had been confirmed. The arguments still needed careful scrutiny before anyone could know whether either was correct.
The Paradox-associated proof was especially hard to assess in its released form. Alpöge’s four-page document was called “garbled” by Andrew Granville, while James Maynard said the proof was then available as incomprehensible Lean code. Tristan Buckmaster described it as crazily complicated and said it would take a long time for mathematicians to digest.
The human-written preprint was also issued as a hurried announcement, before the planned year of polishing and wider discussion. That does not show it is wrong; it means the draft still needs attention. Maynard suspected Alpöge’s proof would be correct, but suspicion is not confirmation. Independent checking and understanding remain essential.
What is the Riemann hypothesis, in terms of the zeroes of the zeta function?
The Riemann hypothesis is a conjecture about where the zeroes of the Riemann zeta function occur. A zero is a value z for which the function equals zero: ζ(z) = 0. The article describes these values as points on a plane, rather than ordinary numbers on a line.
Riemann conjectured that, apart from some obvious but uninteresting cases, all the relevant zeroes lie on a particular vertical line. The article does not give that line’s coordinate; in standard mathematical notation, the conjecture places the nontrivial zeroes on the line with real part one-half. It is the shared location of all those zeroes that the hypothesis predicts.
This matters because the article says the zeta function encodes properties of how prime numbers are distributed. Riemann showed that the distribution of primes is encoded in its zeroes. If the conjecture holds, the zeroes’ precise alignment would support a powerful description of how primes are scattered among whole numbers.
How large is the Millennium Prize challenge: how many problems are there, and what reward does the institute offer for a solution?
The Millennium Prize Problems are a set of seven difficult mathematical challenges selected by the Clay Mathematics Institute at the turn of the twenty-first century. The Riemann hypothesis is one of the seven. The article describes these problems as some of the most challenging in mathematics.
The scale of the reward is substantial: the institute offers US$1 million to anyone who devises a solution to one of the problems. That means the prize is attached to each problem, rather than being a single award shared across the entire set. The article also notes that the institute has its scientific headquarters in Oxford, UK.
A valid, confirmed solution to the Riemann hypothesis would therefore address both a landmark mathematical question and one of the institute’s prize challenges. But a claim alone does not earn confirmation: the proof must stand up to mathematical checking. The article reports two claims, but neither had yet been confirmed.
If the Riemann hypothesis is true, what would it tell mathematicians about the distribution of prime numbers?
Prime numbers appear among the whole numbers in an irregular pattern. The article explains that the Riemann hypothesis matters because it connects that distribution to the zeroes of the zeta function. Riemann showed that the distribution of primes is encoded in those zeroes, making their locations mathematically important.
The hypothesis predicts that, apart from obvious but uninteresting cases, the zeroes all lie on one vertical line. If that prediction is true, Riemann showed it would imply that primes are scattered among whole numbers in a way that is essentially as random as it could be. This is the key consequence described in the article.
That does not mean prime numbers are literally chosen at random. Rather, the hypothesis would tell mathematicians that their distribution has a particular degree of irregularity. The article presents this as a major reason the problem is important. Whether the two newly announced proofs establish the hypothesis remains unresolved while mathematicians examine them.
Why can a computer proof written in Lean still take mathematicians a long time to understand?
Lean is a programming language used to write formal mathematical proofs. Such code can express mathematical steps precisely, but a computer-readable proof is not automatically an easy explanation for a human reader. Checking whether formal steps follow and understanding the ideas behind them are related but distinct tasks.
The article says Alpöge’s proof was then in the form of incomprehensible Lean code. Tristan Buckmaster called it crazily complicated and said it would take a long time for mathematicians to digest. James Maynard also agreed that understanding the corresponding argument would take time. The article does not say the proof had already been independently verified.
Mathematicians therefore still need to examine the reasoning, clarify how the steps fit together, and develop a human understanding of the argument. That process matters especially for an extraordinary claim about a Millennium Prize Problem. Maynard suspected the proof would turn out to be correct, but the article treats confirmation and understanding as work still ahead.
What are prime numbers, and how can the zeta function help reveal patterns in where they occur?
A prime number is a whole number greater than one that has no positive divisors except one and itself. This standard definition is background knowledge; the article focuses instead on how primes are distributed among whole numbers. Primes are not evenly spaced, which makes patterns in their occurrence a deep mathematical question.
The article says the Riemann zeta function encodes the randomness properties of prime numbers. Its zeroes are values z for which ζ(z) = 0, and Riemann showed that the distribution of primes is encoded in those zeroes. In other words, studying where the zeroes lie can reveal information about how primes are scattered.
The Riemann hypothesis predicts that all nontrivial zeroes lie on one vertical line. If true, Riemann showed this would imply that primes are distributed in a way that is essentially as random as possible. The zeta function therefore provides a bridge between the locations of its zeroes and the pattern of prime numbers.
This brief was written by AI from the original reporting and checked by other models. Names, figures and quotes come from the source; read it for full context.
Read more in the JupiteX app
Pulse is free. New stories every 4 hours, each one broken into the questions that explain it.
Or read more news on the web