Integrating chatbots (ChatGPT) in the process of manuscript writing and proposing a roadmap for their future adoption

Abstract

Integrating chatbots (ChatGPT) in the process of manuscript writing and proposing a roadmap for their future adoptionMuhammad Ali1,*, Nazir Ahmad2, Hussein A Younus31 Department of Life Sciences, School of Science, University of Management and Technology (UMT), Johar Town, Lahore 54770, Pakistan. muhammad_ali@umt.edu.pk; ali4982@gmail.com2 Department of Chemistry, Government College University Lahore, 54000, Pakistan. dr.nazirahmad@gcu.edu.pk3 Nanotechnology Research Centre, Sultan Qaboos University, PO Box 17, PC 123, Al-Khoud,Oman. hay00@fayoum.edu.egArtificial intelligence (AI) language chatbots overwhelmed the scientific community these days, specifically the ChatGPT1 . From writing assignments to the original manuscripts, it influenced to an extent that it was listed as one of the authors at few instances2,3raising debate among the scientific community. Before AI language bots, language editing tools were commonly used by the scientific community, however, chatbots are considered superior in context with the speed, refined language, and the so-called contribution. However, there are some serious concerns of the global scientific community regarding the chatbots’ judicial use.Although ChatGPT is a promising tool assisting scientists encompassing relevant literature / data and writing manuscripts in an impressive way, it is not without the associated threats, like generation oferroneous citations.4 Also, the output may be manipulated supporting arguments with tailored efforts, the text-output generated, therefore, needs critical review and consultation(s) / verification(s) with the original literature.5Looking into the recent updates, and to allow its judicial use, the following measures may be taken: (a ) authorship for the language chatbots should be discouraged as these tools cannot be held accountable for any statement or any ethical breach, (b ) it would, in similar way, not appropriate to acknowledge the tool in the ‘acknowledgment section’ as the more appropriate was would be to cite it, instead, and (c ) the most importantly, there are tools developed to detect text generated with language chatbots, at least GPTZero6 . And that the experience with ChatGPT has raised concerns over both the health of the generated content and citations, there is a dire need to re-enforce measures. One way to deal with it is to get disclaimer or the more appropriate would be to have a cutoff value of using ChatGPT in a manuscript. Just like a cutoff for similarity index/plagiarism (i.e. , 19%), for instance. Similar cutoff for the judicial use language chatbots be devised. This may, on one hand, allow its judicial use and, on the other side, will allow the reader the extent to which the manuscript under study be relied upon.

Ähnliche Arbeiten