Here is a 3-paragraph summary for mathopen.com:
Mathematician Terence Tao has proposed an experimental challenge that sits at the intersection of artificial intelligence and number theory. The challenge is inspired by recent mathematical activity around a famous unsolved problem involving so-called Landau-Siegel zeroes, which are hypothetical zeroes of L-functions that, if they exist, would have major consequences for our understanding of prime numbers. Rather than using AI to replace human mathematical reasoning, Tao frames this as an opportunity to explore how AI tools can actively support and sharpen human mathematical thinking.
The specific challenge centers on using "tenth discriminant bounds" as a potential avenue for ruling out Siegel zeroes. Discriminant bounds are technical estimates in algebraic number theory, and pushing them to higher precision opens up new lines of argument that mathematicians have not yet fully explored. The hope is that AI assistance might help researchers navigate the complex web of estimates and inequalities involved in such arguments, potentially uncovering approaches that human intuition alone might overlook.
This initiative reflects a broader vision Tao has been developing about what he calls "Math 2.0," a framework for thinking about how AI and human mathematicians can collaborate productively. Rather than viewing AI as a tool that simply executes calculations, the goal is to use it as a creative partner that helps expand the frontier of mathematical discovery. The challenge is open in spirit, inviting the mathematical community to experiment with these ideas and share what they find.