What the Librarians are Currently Reading on Artificial Intelligence
Artificial intelligence is rapidly reshaping research, publishing, and clinical practice—and raising important questions about accuracy, accountability, and human judgment. This selection highlights articles librarians are reading on AI’s growing influence in scholarly communication and healthcare.
The authors explore the terminology of AI hallucinations through the lens of social stigma. The phrase "AI misinformation" is suggested as a more appropriate label for the problem as AI becomes integrated into medicine.
The authors posit that research evaluation reforms should be organized around 'distinctly human contributions' which may not be well described by metrics.
Large language models may be generating non-existent references. This article explores the issue of fake citations. "Hallucinated citations that make it into the academic literature can slow down and confuse other researchers’ efforts, and lead to false conclusions. Such errors can also create distrust in science,"" says Weber-Boer. Additionally, some publishers say that hallucinated references can be grounds for rejecting a publication and may not allow for resubmission with corrections.