Oliver Braunling, Professor at FH Dortmund.
Today (7. Oct 2026) a preprint of Jonathan Pila and Jacob Tsimerman appeared on the arxiv (https://arxiv.org/abs/2610.09714), with the AI disclosure.
No AI was used in the research or writing of this paper.
It immediately leads to the question: is it possible that researchers like Jacob Tsimerman can compartmentalize their brains to such an extent that one part has been a heavy user of AI for years, while another part of the brain pursues only AI-unrelated thoughts.
While it sounds quite unbelievable to me that humans could isolate their thinking so dramatically, literally like plunging their brains into sandboxes – I also feel that the maths community as a whole will not be able to separate these worlds. How would we? Right now, we see a transitional phase, where the concept of an “AI-free mathematician” might still exist. But we will read papers all the time whose AI disclosures might mislead us (a drastic example for such a disclosure comes to mind right away), and we learn from these papers, oftentimes subconsciously. After a year of reading, refereeing, the daily arxiv briefs – Who will be able to label their knowledge according to its provenance?
And we discuss with others, at the blackboards, over coffee, by e-mail – Many of these exchanges will be shaped by things that some people have learned through AI, they mix, intermingle, and one day there is no other way but to call it a human-AI fusion cuisine.
I would argue that the “AI Disclosure” is only an instrument of very temporary meaningfulness (a view that has been expressed before in other posts). It pretends to measure something that, already today, is not measurable. It is an act of voluntary self-deception.
Or, to put it more mildy: I am unable to extract any meaningful information from Pila and Tsimerman’s AI Disclosure “No AI was used in the research or writing of this paper.”. I know an \emptyset when I see it.
What I, however, can extract from it, is the provocation. All the countless young mathematicians who have just been told that the future of research is the machine, whose belief in their own brainpower has been shattered. Some say, their meaning of life has been evaporated. Some have already left maths, and many are considering to leave. Just now. Perhaps today. This hour. Perhaps precisely in the hour when some other people chose to upload their preprint, where they claim that THEY can still do AI-free research. Exactly the gift they took away from the next generation.
Personally, I have had many conversations with AI and have learned a lot of interesting maths. Yet this text is written entirely by me, no AI use whatsoever. Find out for yourselves whether you consider my text AI-free or AI-shaped, it’s a Schrödingerian cat of sorts at best.
With many, I share the worry that our community will be divided, friends become enemies already today – Let us not have this divide over AI infused ideas or disclosures, but in regards to who acts responsibly towards the next generation of mathematicians. This will be a divide as well, because it makes it necessary to be clear about who is acting absolutely irresponsibly. They are dancing in a rain of tears.
Received 7 October 2026.
Add to the discussion