A team of six mathematicians led by Fields Medalist Maryna Viazovska spent two years breaking down her celebrated eight‑dimensional sphere‑packing proof into formal steps using the Lean system. Graduate student Sidharth Hariharan at Carnegie Mellon rushed to his adviser after receiving an email that the effort had been completed by an AI system. The collaboration was intended to deepen understanding of the underlying geometry.

The AI, called Gauss, was built by California start‑up Math, Inc.. It ingested the group’s roadmap and produced a complete formalization in five days, outrunning the human team’s months‑long schedule. The rapid finish sparked an outcry among researchers, who worry that such speed threatens traditional training, authorship credit, and the market for paid mathematical consulting, and could reshape academic publishing norms.

OpenAI and DeepMind have recently publicized similar breakthroughs, disproving an 80‑year‑old theorem and solving nine more, fueling a competitive race for mathematical AI. A global declaration signed by leading scholars urges transparency and safeguards against monopolizing proof generation. The episode shows investors that AI‑driven reasoning tools can command valuations, but also that the field may face regulatory scrutiny. Investors eye IPO prospects for such firms.