What you'll learn:
- Insight into using AI for scientific publications.
- How AI could help mitigate challenges.
- The potential impact of applying the technology.
Artificial intelligence has become prominent in nearly every industry, from manufacturing and automotive to agriculture and defense and beyond. According to Howard Bauchner, MD at the Department of Pediatrics, Boston University Chobanian & Avedisian School of Medicine, AI will have a hand in scientific publications.
In a recent publication entitled "Artificial Intelligence and the future of scientific publication," Bauchner details how AI's role in medicine is still evolving. Nonetheless, he indicates how it will help evaluate medical images (radiographs, X-rays, videos, etc.), prepare discharge summaries and consultive evaluations, and even assist in diagnosis.
AI, according to Bauchner, will also play a role in scientific publications, specifically peer reviews and drafting manuscripts, and could potentially help editors increase the influence of their journals. He argues that in the coming years, AI will transform the writing of scientific manuscripts, assist in reviewing them, and help editors select the most impactful papers. This notion isn't far off, considering the amount of information, both accurate and false, ChatGPT has produced over the last few years in the science and engineering fields.
Using AI to Generate Initial Paper Review Scores?
In the paper, Bauchner goes on to explain how AI could be used by editors, noting that identifying enough peer reviewers is getting increasingly difficult. Thus, editors could use the technology to provide an initial score.
Articles with good scores could then be sent for external peer reviews. Those with bad scores could still be considered for publication after reviewing it or even have the authors revise the paper. As AI acquires the capability to predict citations, which influences journals' impact factor, the question of whether to use that information becomes a factor.
"First, editors should establish a vision for their journal—what is its mission—and is an individual article consistent with the mission and 'in scope.' Second, editors need to carefully consider the role of value-added pieces. How do they enhance the value of the journal? Third, editors need to maximize the reach of their journals, particularly in social media. Journals are communication networks," explained Bauchner.
"Fourth, editors need to understand the meaning of open science, including open peer review, data-sharing, and open access. After an editor has thought through these issues, then yes, having AI assist in determining how much an article would be cited—assuming the results of the study are valid and simply not meant to attract attention—is reasonable."
Bauchner stated that AI won't replace editors or peer reviewers. Rather, it would provide additional information on the quality of those papers, making the process of scoring manuscripts faster and more objective. It's no longer a question of whether AI will be used for scientific publications, but when it will be used, as drafting quality manuscripts and peer reviews are challenges that investigators, peer reviewers, and editors are currently facing.
