Did you even read the abstract of the TFA? This should not even be a cultural background issue. As an L2 English speaker myself I have never ever thought about throwing a thesaurus onto some established phrase so I can turn “artificial intelligence” into “counterfeit consciousness”, or “deep neural network” into “profound neural organization”. These are deliberate use of fancy words without trying to make sense.

Heck, we got a word for this sort of rampant plagiarism masking on Chinese internet — 洗稿 (manuscript (or blog post)-laundry).

OT: I do appreciate the funny phrase “elite figuring” for HPC. It’s kind of like how they translate things to Anglish.

No one would write "colossal information" instead of "big data" because English isn't their first language.

If you read the article the authors directly address this and use datasets of machine translated articles as controls.

People who don't speak English natively could use machine translation, and people plagiarizing could use machine translation. How do they distinguish (if you don't mind saving me digging into the research)?

Phrases like AI and big data are already pretty well defined in almost every major machine translation set. You'd have to forcefully try to thesaurus your way through to make it do that 99% of the time.

But wouldn't the error rate be the same for both legitimate and plagiarized texts? How do these errors distinguish between the two cases?