In a world of publish or perish, academics will too often turn to our helpful robot friends and ask a chatbot to write the text of a paper for them. Sometimes they don’t check too closely if the resulting text makes any sense.

But there are plenty of low-quality open-access journals that will run any old rubbish if you pay the swingeing page fees. Some of these are published by Elsevier.

Today’s tell is the phrase “vegetative electron microscopy” — an OCR artifact from scanning a paper with “vegetative” in the first column and “electron microscopy” next to it in the second column and reading it as a single phrase.

This phrase is meaningless. It’s obvious smoking gun evidence of a faked paper generated to increase an academic’s publication count.

But it keeps showing up. “Vegetative electron microscopy” is present in about 20 papers on Google Scholar, including one in an Elsevier journal. Elsevier told Retraction Watch they had no problems with the text of the paper containing this glaring tell for LLM spam. You might think this was because the check had cleared long since. [Google Scholar; ScienceDirect, 2024; Retraction Watch]

The phrase was discovered in November 2022 by user “Paralabrax clathratus” on PubPeer. Another user, Alexander Magazinov, found the likely 1959 source paper for the erroneous scan. The phrase is now in Guillaume Cabanac’s Problematic Paper Screener — along with “bosom peril,” “kidney disappointment,” “fake neural organizations,” and “lactose bigotry.” [PubPeer, 2022; PubPeer, 2022; The Conversation]

Pivot to AI

It can't be that stupid, you must be prompting it wrong

https://pivot-to-ai.com