Show HN: UIPalooza – Do AI models have a recognizable design taste?

This question has been bugging me for a while: do different AI models have a different design sense? Does it matter which one I pick?

I built UI Palooza to explore that. I gave models the same four frontend briefs, with enough room to choose their own typography, colors, layouts, and interactions. Then I repeated the runs with different design skills and sample directions.

The public collection has 1,784 designs. You can compare them side by side and open the working websites, not just screenshots. No signup required.

One way to explore it: pick a brief and Baseline A, then compare Vanilla with a design skill. Or keep the skill fixed and look across models.

I’d be interested in which patterns you notice—and what you’d change about the experiment.