A thought leader in the AI/academia space, Allison Papini poses a complicated question: If traditional citation and publishing systems weren’t built with artificial intelligence in mind, how does a person cite AI in papers? This query has led to conflicting practices where some researchers are citing ChatGPT as a co-author, while other journals are prohibiting its use altogether. 

“Using things like ChatGPT is going to make the publication process even more challenging for researchers,” says Papini, assistant director and manager of research and instruction services for Bryant’s Douglas and Judith Krupp Library.  

Across the nation, student and faculty researchers are grappling with how artificial intelligence fits into higher education — both academically and ethically. In the last year alone, Papini has presented on the subject at West Point, the International Association of University Libraries, and other settings; her forthcoming paper, to be published in Educause Review, explores cross-campus approaches to building a generative AI policy.   

Bryant’s librarians are on the front lines of this heated debate. More and more, students, faculty, and staff are looking to AI for their research needs; the Krupp Library staff help them evaluate when and how to use AI for research.  

She notes that one app, Research Rabbit, integrates with reference management software Zotero, so individuals find additional research related to their topic and authors. The idea itself is not particularly novel, says Papini; researchers can do similar things using Google Scholar and EBSCO, which don’t run on AI. 

“The difference is the way this tool makes connections between authors and visualizes your collection of articles,” Papini says. “This could be a significant development for people conducting higher-level research using dozens of articles or even an entire body of research.” 

AI’s use in research includes analyzing large amounts of data, finding keywords for searching, summarizing information for researchers, and improving accessibility to digital archives and other collections. Papini explains that this technology is great for people who aren’t sure how to come up with a research topic or need to get through a lot of information quickly. 

“There’s a ton of potential for AI in research — it’s just a matter of seeing how well it comes together,” Papini says. 

She notes that amidst the technology’s pros, cons are still prevalent — especially with the well-documented issue of AI inventing plausible-sounding references to research articles that don’t exist.  

“It’s good to remember that AI, especially generative AI and large language models, are really good at trying to predict things, but they’re only predictions, not necessarily statements of fact.” 

“This can be frustrating for librarians, faculty, and students alike, which we saw during the spring semester. People will go in circles trying to find an article from a real author in a real journal that the AI invents,” Papini says. “It’s good to remember that AI, especially generative AI and large language models, are really good at trying to predict things, but they’re only predictions, not necessarily statements of fact.” 

On a philosophical level, she notes that using AI to take shortcuts in the research process could affect individuals’ understanding of a topic. She says this is less harmful for faculty who are experts in their field but might be a disservice to students. 

Looking to AI’s future, Papini hopes the technology will make research more accessible for people for whom traditional resources don’t work well. 

“Optical character recognition is a great example of this. It uses machine learning to recognize characters on a page, which then makes it possible to use screen readers, audio transcription, and translators. OCR has come a long way but isn’t foolproof,” Papini says, adding that AI will make literature reviews more comprehensive and make further connections between library collections. 

In the fall, Bryant joins the “Making AI Generative for Higher Education” project through Ithaka S+R, a two-year research project in collaboration with a select group of universities committed to making AI generative for their campus community. 

“We’re looking forward to being part of the conversation and contributing to this important research project,” Papini says. 

To learn more about the implications of AI in the classroom and the research lab, visit library.bryant.edu for a list of recommended readings.