For most of my career, the job was understood narrowly. A designer made things look good and work well, then handed them to engineers who made them real. Visuals and UX, with a wall in the middle.
I think that definition is dying, and I'm glad to watch it go. Here's roughly what I believe instead.
Designers are builders now
The gap between an idea and a working version of it has nearly closed. With AI and a little code, a designer can build the thing, not just describe it. That changes what the role is for. We're no longer decorating a product that someone else assembles. We're shaping the experience directly, hands on the actual material.
I don't see this as a threat to design. It's the most leverage designers have ever had. The person who understands how something should feel can now build it feeling that way, with no translation loss in the handoff.
A builder is a decision-maker
Here's the less comfortable part. If you can build, you are choosing what gets built. You can't hide behind "I just did the mockup" when the mockup is a working prototype with real behaviour in it.
So designers are decision-makers now, whether or not the title admits it. Taste stops being an opinion about pixels and becomes judgment about the product. That's a bigger job, and it carries the weight of being wrong in public. I'll take it.
Rapid prototyping is the new sketch
The fastest way to know whether an idea is any good is to make a cheap version and feel it. Not describe it, not diagram it. Feel it.
Prototyping used to live near the end, a production step. It's moving to the beginning and becoming how we think. A prototype you can hold for an afternoon teaches you more than a week of debate, because it answers the only question that matters: does this actually work, or did it just sound good in a meeting?
Alignment is part of the craft
Good design dies in rooms where three people wanted three different things and nobody noticed until it shipped. I used to treat alignment as overhead, the boring tax between me and the real work. I had it backwards.
Getting people to agree on the problem, early, is design work. It might be the highest-leverage design work there is, because everything downstream inherits it. A beautiful solution to a problem the team never actually agreed on is just expensive waste.
Craft is a mindset, not a tool
No tool will give you craft. Not Figma, not the newest model, not whatever workflow went viral this week. Craft is a standard you decide to hold, then keep holding when it's inconvenient.
That's the catch. Craft is easy to say and hard to do. Everyone nods along that quality matters. Far fewer hold the line on a Friday afternoon, when the thing is good enough to ship and would be better with two more hours. The commitment is the whole game. The talking is free.
In a startup, habits are what survive
I recently wrote about Linear, and the lesson that stuck wasn't a secret. Their craft comes from small, boring habits that nobody skips.
That matters most in a startup, where everything is on fire and motivation is unreliable. You will not feel inspired on the average Tuesday. Taste deserts you under deadline pressure. What's left, when the energy is gone, is whatever you turned into a habit. One small fix a week. Writing the problem down before you build. The reps you do without deciding to. In a place moving this fast, habits aren't a nice-to-have. They're the only kind of consistency that holds.
What I'm holding onto
Designers are builders, and the builders are deciding what gets made. Prototypes are how we think. Alignment is the work, not the obstacle. And craft is a choice you make again every day, mostly through habits dull enough that nobody would post about them.
That's the job now, as I see it. More of a builder, more on the hook, and a lot more fun.