CS SEMINAR

Reclaiming the Writing Process: How AI Interfaces Shape Student Agency and Knowledge Transformation

Speaker
Dr. Hari Subramonyam, Assistant Professor, Graduate School of Education and Computer Science (by courtesy) at Stanford University
Chaired by
Dr Suranga NANAYAKKARA, Associate Professor, School of Computing
suranga@comp.nus.edu.sg

07 May 2025 Wednesday, 11:00 AM to 12:00 PM

i4-01-03 (Innovation 4.0 Seminar Room, next to Nami cafe)

Note: Light refreshments (snacks) will be provided prior to the seminar.

Abstract:
As generative AI becomes embedded in student writing tools, concerns have surfaced around diminished writer agency, superficial engagement, and the erosion of critical thinking. In this talk, I present findings from two studies examining how interface design mediates the role of AI in student writing. I introduce Script&Shift, a layered writing environment that integrates large language models (LLMs) not as co-authors, but as process-aware collaborators across brainstorming, elaboration, structuring, and revision. Across a mixed-methods deployment study and a controlled experiment with 90 undergraduates, we find that Script&Shift significantly outperforms both chat-based and standard writing tools in fostering knowledge transformation and preserving students’ ownership over their writing. I discuss how interface paradigms—such as modular scripting, rhetorical layering, and semantic scaffolding—restructure the cognitive experience of writing, and outline design principles for building AI systems that augment rather than automate students’ intellectual labor.

Bio:
Dr. Hari is a Research Assistant Professor at the Graduate School of Education and Computer Science (by courtesy) at Stanford University. He is also the Ram and Vijay Shriram Faculty Fellow at the Institute for Human-Centered AI (HAI) and a core faculty member of Stanford HCI. His research sits at the intersection of Human-Computer Interaction (HCI) and the Learning Sciences. He studies ways to augment human learning using AI by (1) engaging in cognitively informed design practices, (2) co-designing with learners and educators, and (3) developing transformative AI-enabled learning experiences. Through his research, he also contributes tools and methodologies that prioritize ethical considerations, responsible design practices, and human values when creating AI experiences.
https://haridecoded.com/