It’s been a pretty dark week for me in #mathematics with all the (Open)AI mess playing out. Now that it is here, I have confirmation that I feel zero excitement about the resolution of a millennium problem.

The AMS and EMS have put out these weird corporate statements not saying anything except „let’s all be friends and celebrate“ but does that lift the spirits?

While it is amazing what these models can do, it more and more feels that all this is totally not worth it. So much collateral damage. The despair among early career mathematicians is palpable. The devastating effects on university teaching (which are around for a while) really show their impact now. Lower enrolment numbers and even worse, much less participation and engagement of the students who are enrolled.

I have the feeling that current AI power is enough for a decade of mathematics to come. The companies could just stop training at this point and we would still have extremely powerful tools.

And also we could step by step make AI a public good. It won’t take that long for what we have now to be possible to run by CERN like public infrastructure.