2024-25
THEMES
HCI, AI, COGNITIVE SCIENCE, Human-AI Interaction
RESOURCES
Seeing What Wasn’t There: How Emotionally Charged AI-Generated Media Shape False Memory Undergraduate Researcher, Advisor: Prof. Pattie Maes - Designed A 5-Condition, Sentiment-Controlled AI Pipeline (Runway Gen-3, Adobe Firefly) And Ran A Pilot Study (N=100) Isolating Emotional Valence Vs. Media Modality as a Driver of False Memory Formation. - Built Statistical Pipeline (Welch ANOVA, Games-Howell), Quantifying 2.3x Increase In False Memory Rate For AI-Edited Vs. Control Conditions (18.9% →44.1%, P < .001); Identified Video Modality as Dominant Driver. - Created The Architecture Of The Study That Examines The Impact Of AI-Altered Visuals On False Memories, Which Are Recollections Of Events That Didn't Occur Or Deviate From Reality.



