AI4 Society Seminar | Mina Lee

Abstract: Writing has long served as a finished product and a window into how its writer thinks. Generative AI complicates this, entering the writing process in myriad ways and leaving readers with limited insight into how a text was produced. In this talk, I will present two studies that approach this problem from complementary angles: how AI affects writers’ thinking, and how readers assess writing under uncertainty. First, a preregistered experiment (n=393) shows that the effects of AI use on critical thinking depend on when in the task AI becomes available and how much time people have. Second, interviews with writing instructors (n=28) reveal how they evaluate student writing when uncertain about AI’s role, and how tools ranging from AI detectors to revision history align or conflict with instructors’ needs and values. These studies suggest moving past the binary question of whether AI helps or harms writing, and instead asking how AI’s impact varies across contexts and what matters to the people involved.

‍

‍

Mina Lee is an Assistant Professor in the Computer Science and Data Science Institute at the University of Chicago. Her research examines the evolving relationship between people and AI, with a particular focus on writing, reading, and thinking with AI. She designs and evaluates AI systems, identifies opportunities and risks of AI-assisted writing, reading, and thinking, and measures AI's impact through user studies, controlled experiments, and surveys. Her work appears in both generalist journals (e.g., Nature Human Behavior) and top-tier conferences in HCI (e.g., CHI), NLP (e.g., ACL and NAACL), and ML (e.g., NeurIPS). She was named one of MIT Technology Review's Korean Innovators under 35 in 2022 and received the Young Faculty Award from the Korean Computer Scientists and Engineers Association in America (KOCSEA) in 2025. Her work on human-AI collaborative writing received an Honorable Mention Award at CHI 2022 and a Best Paper Award at CHI 2026 and was featured in various media outlets, including The Economist. She has co-founded and organized multiple workshops, including Intelligent and Interactive Writing Assistants (In2Writing), Human-centered Evaluation and Auditing of Language Models (HEAL), Tools for Thought (TfT), and Broader Impacts of GenAI in Communication at ACL, CHI, and CSCW.

‍