Science Is Drowning in AI Slop

AI causes just an intensification of longstanding problems. The "publish or perish" pressure has compromised research integrity for decades (including manipulated peer reviews, massaged or fabricated results, and ignored conflicts of interest). Cambridge University Press recently warned that "without urgent, sector-wide reform, the global academic publishing ecosystem is at risk of collapsing" (https://www.insidehighered.com/news/faculty-issues/research/...), explicitly attributing these problems to the pervasive publish-or-perish culture.

Ironically, AI even enables scientists to free up some time next to the "publish or perish" nonsense to do some real science, and not wasting all time running for grants and publishing just for the sake of it.

Looking at the Atlantic article, they just seem to focus on the negative findings of the study the article is apparently based on (https://arxiv.org/pdf/2601.13187.pdf). Contrary to the "slop" narrative, the study found that AI-assisted papers often used "more complex language and cite a wider array of sources" and were more likely to surface newer relevant research rather than just recycling old famous citations. This is even "recursive irony": the Atlantic article critiques the "slop" of modern content production while engaging in practices that mirror the very issues it identifies: sensationalism, rapid turnover, and narrative-driven framing over rigorous analysis.