The product engineer

I keep coming back to the idea that one of the more interesting roles of the next few years is someone who can take a product from idea to production on their own: design it, build it, ship it. Not a designer who codes a little, or an engineer with opinions about fonts. Someone fluent in both, with enough business sense to know which problems are worth the trouble.

AI is a big part of why. It already removes a lot of the time and cost between an idea and a working product, and it will keep removing more. A product engineer who uses it well can cover ground that used to take a team.

But there's a catch I don't think goes away: you can only correct what you understand. AI makes it cheap to produce a layout, a component, an architecture. It doesn't make it cheap to know whether any of them are good. That still takes knowing the field: enough design to see what's wrong with a screen, enough engineering to see what's wrong with a system, enough about the business to see what's wrong with the idea.

So that's what I'm learning toward. Not competing with the tools — being the person who can tell them what to do, notice when they're wrong, and add the part they can't: taste, judgment, and a reason for the thing to exist.