AI and mathematical research
Discovery, proof assistance, formal verification, computation, conjecture generation, and collaboration between mathematicians and AI systems.
Research · education · the mathematical profession
A home for events and discussion about advances in artificial intelligence for mathematics—and their effects on science, education, research, and the future of our profession.
About this initiative
Rapid advances in artificial intelligence are beginning to affect how mathematics is discovered, verified, taught, and communicated. This page serves as a continuing home for panel discussions, lectures, and related events examining these developments and their broader consequences for science and education.
The focus extends beyond the capabilities of current AI systems to the choices facing the mathematical community: how research practices may evolve, how students should be trained, which human skills will remain essential, and how these changes may shape the future of the profession.
Continuing themes
Future events can focus on one or more of these interconnected areas.
Discovery, proof assistance, formal verification, computation, conjecture generation, and collaboration between mathematicians and AI systems.
How curricula, graduate education, assessment, exposition, and the skills expected of the next generation of mathematicians may change.
Research careers, the academic job market, mathematical communication, professional norms, and the role of distinctly human judgment and creativity.
Events and archive
The Math and AI series brings together mathematicians, computer scientists, educators, and researchers to examine how artificial intelligence is changing mathematical work. Events range from broad community discussions to focused introductions to emerging tools, methods, and research practices. Recordings, summaries, slides, and related resources will be added when available.
Upcoming events
Upcoming · 1:20 PM ET · Zoom · Registration closed
Speaker: Michael Rothgang (Bonn University)
Digitising or formalising mathematics refers to translating mathematical arguments from natural language (e.g. English) into a computer-readable format: this makes it possible to verify proofs with certainty, often improves mathematical understanding, can transform peer-review and enables new levels of collaboration. In recent years and months, there has been spectacular progress in formalising research-level mathematics in diverse fields. In fact, recent breakthroughs (such as, on the Navier-Stokes equations) have been enabled through formalisation. I will explain what the excitement is about and give an idea of the current state of the art.
To illustrate this point, I will explain the state of the art in differential geometry: there is much more to it than prompting an AI. Already finding the correct general definition is a non-trivial mathematical question. I will survey what is currently formalised, what is being worked on and where you can help. I will also comment on technical challenges of working with manifolds in Lean and how to address them.
Registration is now closed. Zoom access details will be sent to registered participants.
Upcoming · 1:00–2:00 PM ET · Online
Speaker: Jeremy Avigad, Carnegie Mellon University
Part of the Data-Enabled Science Seminar at the University of Houston and organized by Nicolas Charon.
The Zoom link will be posted on the seminar page.
Upcoming · 3:05 PM ET · Lov 231 + Zoom
Speaker: Gigliola Staffilani, MIT
Professor Gigliola Staffilani will speak about the report arising from the Summit on PhD Math Education in the Age of AI, and the broader questions it raises for doctoral education in mathematics.
In person: Lov 231 at Florida State University. A Zoom option will also be available; registration is required for Zoom access.
Past events
Past event
Speaker: Zina Ward (FSU)
Mathematical advances by AI systems over the last six months have prompted worried reflection on the future of mathematical inquiry. Many of the questions mathematicians are discussing deal with topics that philosophers of science have long been interested in, such as scientific change, the nature of mathematical knowledge, how knowledge production is organized, and the (multiple, competing) aims of research. In this talk, I’ll introduce a few philosophical ideas that may be useful for thinking about mathematics’ current predicament, with the aim of continuing the conversation unfolding during this semester’s colloquium series.
Past event · 1:20 PM ET · Zoom
A broad discussion of current AI capabilities, mathematical practice, graduate training, and careers.
Past event · 1:20 PM ET · Zoom
Speaker: Max Weinreich
In this talk, I present the case for total opposition to the use of artificial intelligence in mathematics. I offer proposals for how individuals, departments, journals, and institutions can act in concert to make sure that mathematics survives the coming crisis.
Related paper: arXiv:2608.02859