AI is going to undermine all scientific progress we have made these last couple of centuries. But a handful of techbro oligarchs are going to add some billions to their piles of billions, so its all worth it.

So they found non-existing references in 0.1% of 2.5 million papers examined, and only 0.01% contained more than 2 fake references. Their method also appears to have had a false positive rate for fake references of 7/10.

While the problem is concerning, and growing (fuck AI), 99.9% of papers are ok. We have always known we have to read every paper critically, now there is just an extra thing to look critically at.

In my experience, a way bigger issue is references that don't actually contain or proof the information for which they are referenced.

Note that this is for papers published between 2023 and 2026. No doubt a product of AI.

another AI revolution benefit!

I don't know about the methodology. As an example there was a pretty well known paper in my field that had two different citations. I asked my department chair about it once. He told me that he was actually responsible for it - he had made an error in citing the paper in a high profile report. He said you can see who just copied the citation from him or someone who had. People who actually read that paper copied it correctly.

I have other examples of this kind of thing. The point is that I often look at the cites before I read a paper (partly vanity). If there is an unfamiliar cite in there, particularly if it's an important point, I am going to go look it up. That's the way that science works. The idea that there is a giant citation falsification machine out there is just ridiculous. Anybody who cites something important is going to be found out the moment the rest of us go looking for the source. If you are citing something trivial why bother?