AI, human insights work together to decode prime number mysteries
While AI models did not replace a mathematician in either development, mathematicians have remarked on their ability to find new arguments in the pursuit of cracking old problems
In recent developments, artificial intelligence has been assisting in advancing mathematical breakthroughs, particularly in the study of prime numbers. Prime numbers, which are divisible only by 1 and themselves, play a crucial role in securing digital communications and transactions. The Riemann zeta function, connected to prime numbers, has many non-trivial zeroes, and the Riemann hypothesis posits that all these zeroes lie on a specific line in the complex plane, known as the critical line.
Proving this hypothesis would enhance the ability to predict prime number distribution and potentially expose weaknesses in cryptographic systems.
Anthropic's AI model Claude provided an initial result, showing that at least 41.6% of the non-trivial zeroes lie on the critical line. Jarred Sumner, an Anthropic employee, fed the problem to the model, which subsequently demonstrated that at least 67.2% of the non-trivial zeroes are on the critical line. While mathematicians were able to validate Claude's result, understanding its implications proved challenging.
Separately, mathematicians have long wondered whether there are infinitely many pairs of prime numbers with a fixed gap. In 2013, Yitang Zhang proved that there are infinitely many pairs of prime numbers separated by less than 70,000,000. Subsequent efforts reduced this bound to 4,680, 600 and finally 246, suggesting that there are infinitely many pairs of prime numbers whose gap is at most 246. The twin prime conjecture, a major unsolved problem, would replace 246 with 2.
On August 31, Julia Stadlmann, a U.S. mathematician, published a paper reducing the gap bound from 246 to 240. Axiom Math, a company combining AI, programming languages, and mathematics, reported a further reduction to 212. Axiom Math's team utilized extensive computational experiments and an autonomous theorem prover named AxiomProver to check the resulting proof.
This marks a collaborative effort between humans and AI, where mathematicians developed the argument, engineers manipulated numerical parameters, and the AI system verified the proof. While AI models did not replace mathematicians in either development, they have shown promise in finding new arguments and simplifying complex proofs.
However, many aspects of the mathematical discovery process still require human intervention, such as selecting problems to investigate, interpreting machine-generated arguments, and determining the soundness of mathematical assumptions.
Written by urgent.news from The Hindu - Sci-Tech's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
