Young KWG Panel discussion

Generative AI in Mathematics Research and Teaching

Tuesday 2 April, 11:30 – 12:15

This session consists of three short talks by experts from research fields related to generative AI, and will cover the consequences of the developments in AI for mathematics research, industry mathematics, and teaching.

The speakers for this session are:

• Stefan Buijsman (TU Delft) – Formal proofs versus informal proofs: when proving theorems mathematicians approach problems very differently from current AI systems, distinguishing between trivial and interesting proofs, explanatory and non-explanatory proofs and more. With AI unable to replicate our understanding of a situation where does that leave the interaction with these systems?

• Antske Fokkens (VU Amsterdam) – In this talk, I will walk through the basics on how large language models work. I will focus on why they are so good at certain things and bad at others.

• Wouter Kool (ORTEC) – Learning to optimize or learning what to optimize? Should AI generate a solution to your problem or can you use AI to define your problem? I will discuss the trade-offs using machine learning for optimization, in academia and in practice.

After the lectures, there will be a panel discussion with the speakers. What is the impact of generative AI on our daily work as mathematicians?

This session is organized by Young KWG, a new section of KWG which aims to attract and support young mathematicians in the early stages of their careers.