Here's how ChatGPT quietly breaks UX expectations
· 1 min read · AI UX, Product Design, Design Systems
ChatGPT has evolved – mostly under the hood. But on the surface, things started to diverge.
The session history treats Custom GPTs and regular chats the same. Same UI component. Same behaviour. Same three-dot menu. It works as expected… until it doesn't. Try moving a Custom GPT into a project folder. That menu option? Gone.
No feedback. No explanation. Just a quiet disconnect – the kind that happens when interfaces smooth over complexity instead of acknowledging it. A side effect of prioritising consistency over clarity.
Key takeaways:
- As functionality grows, clarity becomes more important.
- I'm all for consistency – but not at the expense of good UX.
- Sometimes, consistency can be misleading.
- Sometimes, good UX simply means saying: "This works differently."
So how can ChatGPT fix this?
- Enable the same action across both types – a technical fix.
- Or add a small visual cue to signal the difference – a visual fix.
Design for simplicity – not the illusion of it.
More writing
Where does intelligence belong?
Most AI briefs say “add an assistant”. The better question is where AI should help, and how much. Five rules for designing AI into a complex product, with an example from KNIME.
AI UXExperience architectureProduct DesignAssistant usage is not the metric. Resolution is.
Teams measure AI features by how often people use them. That rewards the wrong design. Five behaviours that tell you whether an intelligence layer is doing its job, and how to instrument them from the first release.
AI UXMetricsProduct Design