Please join us for a UW Data Science Seminar featuring University of Michigan School of Information Lecturer and Research Investigator Elle O’Brien on Tuesday, October 14th from 4:30 to 5:20 p.m. PT. The seminar will be held in IEB G109.

“The Future of Scientific Code in the Era of Generative AI”

Abstract: This talk will present findings from three empirical studies involving interviews, user log reviews and surveys that examine how scientists use generative AI coding tools. We find that scientists overwhelmingly use general-purpose chat tools like ChatGPT over dedicated developer tools, which may have consequences for the cognitive load involved in reviewing generated code. We also find that most scientists work in multiple programming languages, and use generative AI tools to navigate less-familiar languages and libraries. Finally, we observe that perceived productivity gains with AI support are inversely related to both programming experience and use of software development practices like testing and version control. Given that scientific code is already undertested and scientists are undertrained as developers, increased reliance on AI-generated code may strain existing code quality control mechanisms in science. We will discuss implications of these findings for research integrity broadly.

Biography: Elle O’Brien, Ph.D. is lecturer and research investigator at the University of Michigan School of Information, where she teaches graduate courses about statistics and data science. Her research program asks how scientists who program use large language models as coding assistants and how they validate generated code. Previously, she studied auditory neuroscience using mathematical models of sensory perception at the University of Washington and worked as a developer advocate for open-source tools like Data Version Control.

This seminar will also be available at this time on Zoom if you cannot attend in person.

The 2025-2026 seminars will be held in person, and are free and open to the public.