Everyone's an AI Designer Now. That's the Problem.

Everyone's an AI designer now. That's kind of the problem.

Put "AI" in your title, spin up a prototype in whatever tool is trending this month, push some code with an agent's help — sure, anyone can do that now. It's not the differentiator it looks like.

Here's what actually is: context.

I'm in my mid-40s. Twenty years sitting across product, design and engineering — not handing off specs, working through the trade-offs with the people building it. Founded my own agency. Built startups alongside founders, owning strategy and roadmap as much as the design work. That's two decades of problems, teams and personalities I can pattern-match against, and it's not something you can shortcut with a better prompt.

I don't say that to gatekeep it. Plenty of designers in their 20s and 30s are doing extraordinary work, and ten years in is real experience too. But there's a version of this moment that quietly discounts what people who've made the mistakes already carry — and I think that's backwards. Call us designosaurs if you want. Knowing what to do, and just as importantly what not to do, is the whole game when you're the one directing an agent instead of doing every keystroke yourself.

Because that's the actual job now. An idea walks in as a bet. You prioritise it against everything else competing for the same time. AI does the synthesis, the research, the first pass at the data. You prototype, and then — the part that never shows up in the demo — you cut it back to what the problem actually needs. Then you ship it: real code, in front of real customers, with tests and guardrails and an agent-reviewed PR pipeline doing the boring parts so you don't have to.

Writing syntax was never the bottleneck people thought it was. Understanding the customer well enough to direct an agent inside the right guardrails — that's what's left, and it still needs a designer.

Delivery got fast. Adoption and retention didn't get any easier — if anything they're the harder problem now. Which, funny enough, is just agile doing what it always promised: tight loops, fast feedback, iteration that doesn't wait for a quarter to prove itself. We just do it daily now instead of talking about doing it.

That's what I mean by AI product designer. Not the tools. The judgment that decides what they're pointed at.