Most designers working on AI products are focused on the wrong layer. They're refining interfaces while the actual experience: how the model responds, what it prioritizes, where it draws lines is being decided in engineering meetings they're not in. System prompts are UX writing. Data decisions are research decisions. Evaluation frameworks are definitions of quality. These have always been design's domain. The tools are just different now.

This workshop closes the gap between the UX skills designers already have and the AI-specific materials they haven't been taught to use. Participants will work hands-on with system prompts, data auditing, and evaluation frameworks, learning to shape what a model does, not just how its output is displayed. The shift from designing screens to designing behavior is the most significant expansion of design's scope in a generation. This workshop is where that shift becomes practical.