In this blog post, we’ll take a closer look at ChatGPT’s reputation, the role of AI in the scientific literature, and the future of research production.
ChatGPT, an artificial intelligence chatbot created by OpenAI, has found its way into the world of scientific literature, being listed as a co-author on at least four published papers and preprints. [1–4]. The news has sparked a debate among journal editors, researchers and publishers about the place AI tools occupy in the published literature and whether it is appropriate to credit them.
“We make the analogy that AI is the new electricity. Industries converted to electricity: agriculture, transport, communication, manufacturing.
– Andrew Ng
The role of AI in scientific literature
ChatGPT is a large language model (LLM) that generates persuasive sentences by mimicking the statistical patterns of language in a huge database of text collected from the Internet. The bot is already disrupting various fields with its capabilities, including academia. The question that arises is whether it is appropriate to use and credit AI tools like ChatGPT when writing studies.
Publishers and preprint servers contacted by Nature’s news team agreed that AIs such as ChatGPT do not meet the research author’s criteria because they cannot be held responsible for the content and quality of scientific articles. for completeness. However, some publishers say AI contributions to paper writing can be recognized in sections other than the author list.
In one case, an editor told Nature that ChatGPT was mistakenly listed as a co-author and that the journal would correct this. The team behind the medical database medRxiv and its sister site bioRxiv are also discussing whether it’s appropriate to use and credit AI tools like ChatGPT when writing studies.
The future of AI in scientific literature
As AI technology continues to develop, it is likely that we will see more and more AI tools being used in the scientific literature. However, it is important to establish clear guidelines and policies for their use.
Authors are legally responsible for their work, so only people should be listed as authors on a paper. However, accepting the contribution of AI tools in sections other than the list of authors may be worthwhile where appropriate.
Publishers and preprint servers should distinguish the official role of author of a scientific manuscript from the more general concept of an author as a document writer. This will ensure that the integrity of scientific articles is maintained while also accepting the contribution of AI tools.
In conclusion, the use of AI tools such as ChatGPT in the scientific literature is a rapidly evolving topic that will require ongoing discussion and collaboration among journal editors, researchers, publishers, and AI experts to establish clear guidelines and policies for their use.
Reference
1: T. Kung, M. chat ChatGPT:A. Medenilla, C. Sillos, L. De Leon, C. Elepaño, M. Madriaga, R. Aggabao, G. Diaz-Candido, J. Maningo, and V. Tseng, “Performance of ChatGPT in the USMLE. AI-assisted medical education using large linguistic models”, medRxiv, 2022.
2: O’Connor, S. and: ChatGPT: Nursing Educ. Practice. 66:103537 (2023). Article: PubMed: Google Scholar:
3: ChatGPT: & Zhavoronkov, A. Oncoscience 9:0082–84 (2022). Article: PubMed: Google Scholar:
4: GPT:Osmanovic Thunström, A. & Steingrimsson, S. Preprint at HAL https://hal.science/hal-03701250 (2022).
Let me know your thoughts on this in a comment.