“Of course. Here is a possible introduction to your subject.” This is the sentence which, until a few hours ago, was read in the summary of a scientific article published in the scientific journal Surfaces and interfaces, by Elsevier, a Dutch publishing house specializing in the publication of medical and scientific publications. The sentence leaves no doubt: the authors of the article used ChatGPT or another AI-based language model to write the introduction. That doesn't mean it's wrong. But among the rules that researchers must follow to publish at Elsevier is that of explicitly indicating whether artificial intelligence tools were used in the preparation of the article. Furthermore, the presence of the sentence in the first line of the summary raises doubts about the (apparently mediocre) thoroughness adopted in peer review of the article. Furthermore, it highlights the pressure that researchers are under, often forced to publish numerous articles per year.
“We are investigating to understand how this happened”
There peer review it is the process by which researchers analyze the work of others to ensure that it has been carried out correctly, guaranteeing the reliability of scientific publications, distinguishing them from less credible ones. In the case of Elsevier, “our policies make it clear that I Great language model (LLM) may be used in writing articles provided it is declared by the authors at the time of submission. We are reviewing this article and are in discussions with the editorial team and authors,” the editor wrote in a post on X (formerly Twitter), regarding the Zhang et al article. fruit of the work of researchers from Peking University and Hangzhou University entitled: The three-dimensional porous mesh structure of Cu-based organometallic-cellulose separator improves the electrochemical performance of lithium metal anode batteries..
Our policies clearly state that LLMs may be used in the writing of articles provided they are declared by the authors upon submission.
We are reviewing this article and are in discussions with the editorial team and authors.
–Elsevier (@ElsevierConnect) March 14, 2024
Publish or perish
Furthermore, as also pointed out by science communicator Ruggero Rollini, the oversight in the introduction could be a symptom of the phenomenon known as publish or publish. Three words which describe a system of pressures, incentives and disincentives to which researchers are often subjected and which push them to publish as many articles as possible, even to the detriment of the quality of the latter. Concretely, in order to obtain citations and prestige, universities, research centers and scientific journals can push researchers to continually write articles, hoping that the volume of publications will increase their weight in the academic community. The other side of the coin is the focus shifting from actual research to writing. Even if there would not be enough content to justify publishing a new article.