Exploring AI as a Catalyst for Cross-Disciplinary Learning
April 21, 2025
Computer Science student McKayl Baily, Biology student Nagashree Avabhrath, Biology Professor Mark Grimes, and Computer Science Assistant Professor Lucy Williams used a 2024 Flagship Fund award to explore how AI could help bridge the learning gap between academic disciplines.
Appreciating that interesting things happen at the intersection of disciplines, UM Biology professor Mark Grimes and UM Computer Science assistant professor Lucy Williams proposed a Flagship Fund project last year to use AI to assist in bridging the educational gap between the students in their disciplines. A graduate student in cell biology, Nagashree Avabhrath, would be paired with a McKayl Bailey, a computer science student, to solve problems associated with writing code to analyze biological data. The pilot project launched last summer and has been remarkably successful.
Avabhrath and Bailey began with a set of functional but loosely organized research code related to cell signaling in cancer. With guidance from Grimes and Williams, they used Perplexity AI and other tools to clean up, annotate, and transform the code into a well-documented R package. Along the way, they experimented with asking AI to explain biological concepts or coding practices, depending on their own expertise gaps.
Williams says that a challenge of interdisciplinary research is the immense background you need to make progress. The project was a way to explore whether AI could help with that problem.
"AI is a nice tool to make interdisciplinary collaborations feel better and be more productive,” says Williams. “It’s an additional resource, another place people can go to communicate across disciplines.”
“AI is like having an extra explainer,” adds Grimes. “We’re trying to define where it's useful.”
Avabhrath expressed initial hesitation about using AI. “I feel like I become a little bit lazier," she says. "It’s the guilt factor.” The experience has helped her appreciate that AI is good at metaphors and analogies that help teaching and understanding.
The second phase of the project was to take what was learned from Avabhrath's and Bailey's summer work and apply best practices in Dr. Williams’ computational biology course offered in fall 2024, and to continue to observe if students were able to use AI to push their course projects further than they could by simply relying on their own knowledge or skills.
The team hopes to gather feedback and share their findings with faculty across campus and beyond. They have already presented at a Symposium on AI in Research and Education at 猎奇重口 Tech in Butte and the Western Regional IDeA Conference, Use of Data Science in Research session, a regional conference for biomedical scientists in Anchorage, Alaska.
The ultimate goal of the Flagship Fund supported project has been to better understand whether, and how, AI can help students learn across disciplinary boundaries—and where its limits lie. “Without the seed money we wouldn’t have been able to do the project,” says Grimes.