New AI Tool Gives Design Students Instant Feedback on Their Work

Design school has a famous ritual called the "crit" — you pin your work on a wall, a teacher and classmates tear it apart, and you go back and fix it. It's one of the most valuable parts of a design education, and also one of the slowest. A student might get detailed feedback once a week. A researcher named Nizam Kadir has built a tool that tries to make that loop faster.

The tool, called Critsly, is an AI workspace where students upload their work-in-progress on a visual board, write down what they were trying to achieve, and then get automated critique from several different perspectives — think "what would a user say?" or "what would an engineer say?" It then pushes students toward a concrete action plan for revisions. A companion mode called StudioCrit adds classroom features: student roles, teacher dashboards, and exportable evidence of who did what. The pitch is simple: more feedback, more often, with a record of whether students actually acted on it.

But here's the important part. The report doesn't prove any of this helps anyone learn. All the testing was done with software simulations — three pretend studio scenarios, plus a rehearsal with 50 throwaway learner accounts. That produced 109 pieces of logged "evidence," and the system sorted most of them into higher-level thinking categories. A later stress test logged 50 completed sessions with no crashes and correctly blocked students from seeing teacher analytics. That's a report card on the plumbing, not on the teaching.

The author is refreshingly upfront about this: no real students, no measured learning gains, and the AI's automatic classifications haven't been checked for accuracy or consistency. So think of this as a promising prototype with a clean audit trail, not a proven classroom tool. What it does show is that AI critique tools are moving from novelty toward something schools might actually buy — and that the hard question, whether they help people learn, is still unanswered.