Duplicated images are often a substitute for real data. https://www.newyorker.com/science/elements/how-a-sharp-eyed-...

Yikes. She contacted the journal to report this clear fraud they were perpetuating and received no response for 6 months. Yet people equate “peer reviewed” and “absolute fact”.

If you want to hear back from any organization, you’ve got to write it on lawyer letterhead.

Which indeed doesn't work for journals because publishing wrong research isn't (and shouldn't be) illegal...

I just mean, it’s so easy to ignore inquiries and I’ve found that if you ever get ignored and need to be heard, a lawyer’s letterhead attracts attention.

I agree with the legal issue you put forth.

Yes, but the social engineering aspect of that trick might still work.

Peer review means "has no blatantly obvious methodological issues" at best, and not even that a decent amount of the time. A lot of people take it to mean that the paper's results/conclusions are accurate.

Peer review standards also differ between fields and journals.

I think this demonstrates a hypothesis: The institution is failing, the technology which exploits the failure is close to irrelevant.

> I think this demonstrates a hypothesis: The institution is failing

This claim needs to be more specific to be meaningful: How is the 'institution' defined? What evidence do we have that it's performing worse than before, or not functioning (producing valuable scientific knowledge)?

I think when lay people say "peer reviewed research" they mean it as a metanym for "the scientific consensus as established by empirical research and vigorous debate," which isn't absolute fact to be sure, but is about as sturdy a foundation as you can hope for. I think I know more about peer review and retractions than most people - and I know next to nothing.

> She found that it contained text plagiarized from eighteen uncredited sources, which she categorized using color-coded highlighting.

That kind of plagiarism is a sign that AI wasn't used, since once tokenized the same concepts are seldom rephrased verbatim. The same problem applies to problematic images. Regenerated-with-alternations is harder to detect than Photoshop. Things will get interesting when you can cheaply generate middle-tier quality scientific papers with a short AI prompt.