John Douglas

PhD thesis:

Explainable Model Adaptation in Extended Reality: Personalizing Trust Through Adaptive Interfaces

This research investigates how personalized Explainable AI (XAI) can build calibrated trust in adaptive Extended Reality (XR) systems. As XR technologies integrate AI models that dynamically modify interface behavior (e.g., repositioning UI elements, adjusting information density, or reconfiguring layouts) users face a critical challenge: they must trust black-box models making real-time decisions about their immersive environment without understanding why changes occur. This opacity can erode trust, particularly when adaptations feel intrusive or unpredictable, and conflicts with emerging legal requirements for transparent automated decision-making. Unlike prior XAI research focused on explaining AI decisions, this work addresses why an AI changed its behavior and, critically, when and how those explanations should be delivered to match individual user needs. 

Who or what inspired you to pursue your research interests?

My main inspiration for studying HCI comes from my personal experiences with poorly designed systems which leave me frustrated or annoyed and my experiences with well-designed systems which leave me feeling empowered. Additionally, my sisters, mom, and dad are my inspirations and provide me with motivation to keep designing human-centric systems that are useful beyond a lab environment.