Can modern scholarship escape AI?
I wrote a paper ...
Is it possible to be an academic, a scientist, a scholar, in 2026, and not have AI impact your work in some way?
And, even more importantly for those scholars grappling with “AI Use” statements when they submit papers to journals and preprint platforms, how do you convey your use while retaining your academic dignity?
To explore this I flexed my considerable academic prowess and wrote a paper which was so radical that even arXiv rejected it!
(The PDF can be downloaded here)
OK, so maybe “paper” is a bit of a stretch here — and it’s not hard to see why it didn’t pass the arXiv bar (although it did take a couple of weeks for the moderators to come to a decision).1
But the point it makes is a very serious one — and extends to any domain where people are expected to articulate their use of AI clearly and concisely, including in classes being taught by professors grappling with the same challenges in their academic work: AI is now so ubiquitous that it is near-impossible to avoid its use in our professional lives.
Of course, this leaves the question dangling of what this means for academic and intellectual work when, even if you think you’re AI free, you are not.
Way more important than any of this though is that, if you are an academic struggling with what you put in your AI Use statement, you now have a template for this.
You’re welcome 😁
The very considered—and considerate—response from arXiv Support was “Thank you for submitting your work to arXiv. We regret to inform you that arXiv’s moderators have determined that your submission will not be accepted and made public. In this case, our moderators have determined that your submission is a content type that arXiv does not accept.” Despite the joke, they do have standards to maintain!