The Worst‐Case Scenario After <scp>AI</scp> Use in Academic Writing: A Clever User Wins?

Abstract

We have a long-standing interest in paper writing, with the first author having written over 570 PubMed-indexed papers during 45 years of obstetrics-gynaecology practice. We compared human-written manuscripts with those generated by generative artificial intelligence (AI) (like ChatGPT), aiming to characterise ChatGPT-generated writing. Here, for simplicity, we will use ChatGPT as representing generative AI. In this way, clever ChatGPT users, those who master maximising ChatGPT's abilities, will produce many papers. Paper acceptance/rejection may increasingly be decided, or aided, by AI, indicating a considerable shift from human–human dialogue to AI-AI combat in the publication process. We write this not in irony but based on our experiences studying AI. We dislike such a future: for us, papers remain something ‘longed for’ [8]. Yet, this ‘nostalgia’ may soon be outdated. Even so, we maintain a deep, personal attachment to writing on our own, a profoundly meaningful human endeavour. We continue to reflect on ‘how meaningful human writing is’. Although this paper does not provide answers, we presented an unsettling vision of the future, which we hope will make readers reconsider the current ChatGPT use in writing. The authors have nothing to report. The authors declare no conflicts of interest.

Ähnliche Arbeiten