My project challenges a foundational assumption in contemporary digital systems: that observable behaviour is a reliable proxy for human intention. Today’s interfaces treat every click, swipe, gesture, and input as deliberate and equally meaningful. Yet human action is rarely so clean. Reflexive and habitual movements, accidental dismissals, mistyped passwords, and iterative rewrites are all forms of friction we encounter within the current paradigm.

It raises questions I want to explore: Are there better forms of interaction to consider? Better yet, are there better forms of technological understanding? How can we train virtual assistants, recommendation engines, and machine-learning models to care about what people mean, not just what they do? How can we build better extensions of the self?