Arnold Named GAITAR Fellow To Study How AI Can Strengthen Student Learning in Labs
Heidi Opdyke
- Associate Dean of Marketing and Communications, MCS
- Email opdyke@andrew.cmu.edu
- Phone 412-268-9982
Annie Arnold, special faculty in chemistry at Carnegie Mellon University, has received a fellowship through CMU’s Generative Artificial Intelligence Teaching as Research (GAITAR) Initiative to study how — and whether — generative AI can meaningfully support student learning in laboratory courses.
Awarded by CMU’s Eberly Center for Teaching Excellence and Educational Innovation in fall 2024, the GAITAR Fellowship provides $5,000 for instructors to design, implement and assess a teaching innovation that uses generative AI.
Arnold’s project focuses on Advanced Organic Chemistry Laboratory, a course she teaches. The study aims to help students make more intentional choices about how they visually represent scientific data.
“When students make figures for lab reports, they rarely stop to ask what their data is actually showing or who needs to understand it,” Arnold said. “Instead, they default to familiar chart types. I want to know whether structured dialogue with an AI can help them reason more effectively about those choices.”
A 2019 CMU graduate with a Ph.D. in chemistry, Arnold appreciates the importance of precise and effective communication of complex data. In 2015, she won Carnegie Mellon’s second Three Minute Thesis competition, which challenges doctoral students to explain their research using a single slide and just three minutes. Arnold’s presentation described her work developing a biodegradable bone implant made from modified graphene. The implant was designed to degrade over time while encouraging new bone growth rather than remaining permanently in the body.
For her GAITAR project, Arnold will lead an in-class activity in which students use the generative AI tool Claude as a structured practice partner. Students will work through a series of figure-selection scenarios, using AI-guided questioning to evaluate their choices, and then complete all graded work independently without AI support. Arnold will study whether this approach is more effective, equally effective, or less effective than providing the same guidance through written instructions alone.
“The goal is not to have AI do the thinking for students. It’s to give them a chance to practice thinking, with guidance, before they’re evaluated,” Arnold said. “I’m genuinely excited about the question and just as interested in what the data says back — whatever that turns out to be.”
If successful, the approach could be adapted to laboratory courses in other fields that also require students to visually represent data. Because the activity relies on publicly available AI tools and requires no new software, the barrier to adoption for other instructors and institutions is low.
The GAITAR Fellowship is part of a broader university effort to study the real impacts of generative AI on teaching and learning. By framing AI integration as a research endeavor, the initiative helps instructors move beyond speculation and toward evidence-based decisions about when and how these tools are most effective.
Arnold brings both academic and industry experience to the project. She returned to Carnegie Mellon in 2024 to teach and work on BioBind, a biotechnology company she co-founded and where she serves as vice president and chief technology officer. Previously, she worked as a materials scientist at Pacific Northwest National Laboratory and at the National Energy Technology Laboratory.
With CMU’s long-standing leadership in learning science and artificial intelligence, Arnold said the fellowship is part of a larger conversation about the responsible and rigorous use of AI in education.
“This is about understanding where generative AI genuinely helps students learn,” she said, “and where it doesn’t. That distinction matters — for chemistry, and for education more broadly.”
