Below the Article, There’s Now a Stack of Papers

I’m sitting in front of my WordPress backend, the article half-finished, twelve tabs open in my browser. Each one contains a preprint, a paper, or something that passes itself off as one. I’ve been searching and researching, as I have so many times before, and at some point I think: if I’m already pulling all of this together myself, why should every reader have to do the same? That’s the moment when, driven by what AI makes possible and by my own situation, a tool is born from necessity.

Throughout my professional life, I’ve always been drawn to practical application. Grand theories and strategic thinking — others have handled that, and many of them better than I ever could. If you’re in sales or business development, you mostly want things to sell or to serve a purpose. You ask upfront whether something makes strategic sense, of course. But some products simply aren’t strategic. Some exist because they’re fun, and fun requires surprisingly little strategy.

AI is exactly that kind of case. It has thrown everything into complete disarray on LinkedIn and elsewhere. You can use it for the most trivial nonsense — „Please create an image of me looking like a hamster“ — or you can use it the way I’ve since learned from Danny Walter: for research and investigation, in an entirely different way. Both are perfectly legitimate. Only one of the two has genuinely fascinated me, and it wasn’t the hamster.

For me, this is a practical application that’s genuinely enjoyable. I’ve written a number of articles here on this blog, searching deliberately for papers in the process, with AI as a research partner. Then it occurred to me: someone might come along, think „hey, I really like what digitalpaddy wrote here,“ and then wonder whether there’s actually any science behind it. Preprints, studies, something solid. That’s easy enough to explore with AI, and out of that, a small plugin was born.

Here’s how it works: beneath articles that I consider relevant, the plugin searches for matching papers. Anyone can click to display the papers found, dive deeper into the topic, and download documents. No trawling through specialist databases, no twelve open tabs. Whether any papers exist for a given article, I don’t know in advance. If there are none, that’s stated plainly. I find that more honest than any pre-sorted reading list. And if nothing turns up, that too is information.

Because a tool that only runs on my setup offers limited practical value, you can now download it yourselves. The plugin is called alphaxiv-papers; the current version is available here as a ZIP file: Download Plugin. The link stays the same with every new version and always points to the latest release. The code and documentation are on GitHub. Give it a try.

This English version was created with the help of AI (translated from the German original) and may differ slightly from a professional human translation.