Architecting Nonlinear Thought: A Structural Approach to Relay–Discontinuity in Human–AI Interaction

Traditional Human–AI Interaction (HAI) research has primarily emphasized usability, efficiency, and performance metrics. While these frameworks remain essential for task-oriented design, they fail to account for the nonlinear cognitive transformations emerging through sustained engagement with generative AI systems. This paper reconceptualizes HAI as a cognitive architectural environment—a space in which thought is not merely assisted, but structurally reconfigured through mechanisms of relay and discontinuity. We introduce Relay–Discontinuity, defined as a thresholded structural divergence (Δ) that transforms, rather than extends, a cognitive state. Rather than treating discontinuity as friction or failure, we argue that it can be deliberately engineered as a productive force.