Response to the Letter by Matsubara
Masaki Mori
Abstract
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsJewish Identity and Society · Legal case studies and regulations · Diverse Scientific and Economic Studies
We appreciate and agree with the comments by Dr. Matsubara that real-world evidence is needed to establish the benefit versus risk of using artificial intelligence (AI) models in writing assistance ^(1)^. We also agree that the method proposed by Dr. Matsubara would be a good strategy to identify the impact of using AI language models.
Dr. Matsubara also raised an important point about clarifying the use of AI for writing assistance even in the case of permitting the use. We think that whenever using writing assistance with AI, it is important to include the following information (1) which AI was used, (2) what version was used, (3) what prompts were used and (4) which part was assisted by AI. This information is important to ensure that publishers and researchers can identify the cause in case of any problem (such as those caused by hallucinations ^(2)^ and adverse attacks, which are still a frequent problem in AI) ^(3), (4)^.
As Dr Matsubara states, the difficulty of this experiment is in its implementation. We totally agree with his point that which journal will take the initiative will be a challenge. In addition to these aspects, the confidentiality of unpublished articles may become a problem. Most AI models are known to use the input data to further improve the AI ^(5)^. This means that the information may become publicly available through the AI before being published in a journal. However, the impact of this fact may have decreased recently since it is increasingly becoming normal to allow pre-prints before publishing the reviewed article.
All of these discussion supports the lack of definitive evidence on how to regulate the use of AI in writing. Further research in this area is urgently needed for this emerging technology.
Article Information
Conflicts of Interest
None
Disclaimer
Masaki Mori is one of the Associate Editors of JMA Journal and on the journal’s Editorial Staff. He was not involved in the editorial evaluation or decision to accept this article for publication at all.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Matsubara S. A hypothetical experiment: evaluating the real-world impact of artificial intelligence use in academic writing. JMA J. 2025;8(2):662-663.
- 2Walters WH, Wilder EI. Fabrication and errors in the bibliographic citations generated by Chat GPT. Sci Rep. 2023;13(1):14045.37679503 10.1038/s 41598-023-41032-5PMC 10484980 · doi ↗ · pubmed ↗
- 3Yamasawa D, Ozawa H, Goto S. The Importance of Interpretability and Validations of Machine-Learning Models. Circ J. 2024;88(1):157-8.10.1253/circj.CJ-23-085738057101 · doi ↗ · pubmed ↗
- 4Goto S, Mc Guire DK, Goto S. The Future role of high-performance computing in cardiovascular medicine and science -impact of multi-dimensional data analysis. J Atheroscler Thromb. 2022;29(5):559-62.34602525 10.5551/jat.RV 17062 PMC 9135644 · doi ↗ · pubmed ↗
- 5Google GT, Anil R, Borgeaud S, et al. Gemini: a family of highly capable multimodal models. ar Xiv preprint ar Xiv:231211805. [Internet] 2023 [cited 2024 Dec 22]. Available from: https://arxiv.org/pdf/2312.11805