Science Is Drowning in AI Slop

The commenter describes AI-generated papers as an escalating burden in their workplace, where people read research regularly. They say papers increasingly contain mistakes and appear to have been reviewed with AI as well, making it difficult to separate useful findings from low-quality work. In their experience, dozens of papers on a topic may yield only one with good information, after hours spent reading. They compare the resulting overload to alert fatigue in security operations: repeated false positives make it easier to overlook a genuine signal, though evaluating papers takes even more time.

The commenter says the problem has discouraged both reading and writing: colleagues are growing frustrated, and researchers may hesitate to publish when solid results risk being buried. They add that paid academic platforms still carry this material. As a coping strategy, they say readers increasingly rely on researchers they already trust, but that makes it harder for new, credible researchers to gain attention. Their most extreme example is a paper built from an AI-generated questionnaire, answers from six people described as field experts, and AI-generated analysis - a process they found infuriating.