When AI can generate screens in seconds, what should we actually teach UX/UI students?
I'm a professor at a regional university in South Korea who was handed a blank canvas: build a 4-year UX/UI curriculum from scratch, starting with the class of 2026 — students who will graduate into a world where most production tasks are automated.
My answer was to stop teaching tools and start teaching decisions.
In this talk, I'll share the competency-based curriculum I designed and am currently running — one that replaces "make pretty screens" with "define problems, test hypotheses, and own the decision log."
Every semester has one core skill, one signature deliverable, and one clear rubric. AI isn't banned; it's embedded — but students must document how they used it, why they chose (or rejected) its output, and what they'd do differently.
This talk offers a ground-level view of what happens when you rebuild design education for the AI era — not as theory, but as a living experiment running right now in a classroom in Gwangju, South Korea.