AI is great at moving. It is not great at taste.
AI can help you ship faster, write faster, test faster. But knowing what to keep, what to cut, and what feels slightly off still takes judgment.
Here's the honest version: AI is great at getting you moving. It is not great at taste.
It can help you ship faster, write faster, test faster. It can produce ten variations of a button or a paragraph or a layout in less time than it takes to make a coffee. That part is real and useful and changing how I work.
What it can't do — yet, in any reliable way — is tell me which one of those ten is the right one. That part is annoyingly human.
I'm not nostalgic about that. I think it's a feature. The bottleneck used to be production: how fast can you make the thing? Now the bottleneck is judgment: which thing should you have made? AI moves work from "can you make it" to "do you know what good is." That's a different question, and it's the one I've been training students to answer for a decade.
So when I'm working with AI on a real project, I lean on it for everything that's "more, please." More options. More iterations. More copy variations. More layout sketches. The "more" is fast and cheap.
The discrimination is still mine. Which option to ship. What to cut. Why this header treatment but not that one. What feels slightly off in a place I can't immediately name. AI is bad at "slightly off." It's a continuous gradient model trying to behave categorically, and the in-between is where the judgment calls live.
This is also why "just use AI" is a bad design philosophy and "never use AI" is a bad design philosophy. Both miss what's actually shifting. The work isn't disappearing. It's redistributing toward the parts that need a human eye.
Which means: invest in the eye. Build the vocabulary. Practice making the call when the call is hard. Those skills used to feel like luxuries on top of production speed. They are now the entire job.