For about a decade, the smart advice to designers was to stop touching the tools. Move up. Stop pushing pixels. Hand the making to other people and spend your time on strategy, on alignment, on the deck. The career ladder was clear. The higher you went, the less you made anything with your own hands.
I never followed that advice. I am fifteen years into leading and doing design, and I still open Figma. I still write the copy. I still touch the front-end. People told me that was a junior habit I should grow out of. They were wrong, and now the cost of that old advice is finally showing up on the bill.
AI changed the math. The distance between an idea and a working thing used to be long and expensive. You needed a spec, a ticket, a sprint, and a list of other people. Now a designer who can build carries an idea straight to something real and testable, alone, in an afternoon. That is not a small speed-up. It is a different way of working.
What the old advice was really protecting
The advice to stop making was not crazy when it started. Making was slow. A designer who insisted on building everything themselves became a bottleneck. So the org chart split the work. Some people thought. Some people made. The thinkers got promoted, and making got coded as the thing you do before you are senior enough to delegate it.
That split made sense when the cost of building was high. It rewarded people who could direct work they no longer did. It produced a generation of design leaders who are excellent at talking about design and slow at producing it. For years that trade was fine, because the people who made things were cheap and plentiful and the gap in speed was small.
The gap is not small anymore. When one person can carry an idea from sketch to working prototype in an afternoon, the cost of being a pure director goes way up. You are now paying a tax to keep your hands clean. The market is starting to notice that tax, and it is starting to refuse to pay it.
Building is not becoming an engineer
Let me be clear about what I mean by build, because the word scares people. I do not mean you need to become a software engineer. I do not mean you ship production code that scales to millions of users. I mean you make your idea real enough to test and real enough to argue for.
There is a huge difference between a picture of a thing and the thing. A picture of a screen gets nods in a review. A working version gets used, and then it gets argued about, and then it gets better. When you can build, you stop bringing pictures to a fight that real things win. You bring the real thing.
Most of the value is in that gap between a mockup and a working draft. A mockup hides every hard question. The animation looks smooth because it is fake. The empty state never shows up. The error never happens. The moment you build it, all of those questions come due at once, and answering them is the actual design work. The mockup let you skip it. Building makes you do it.
AI is what makes this reachable for designers now. You do not have to memorize a framework to get a working version on the screen. You describe what you want, you correct it, you shape it, and you keep going until it behaves the way you pictured. The taste is still yours. The tool just removed the months of learning that used to stand between your taste and a real artifact.
I built a company this way
I did not just write this down as a theory. I built Story Genie as the proof. It is a service where AI agents write and illustrate personalized hardcover children’s books, with humans supervising the quality. A parent answers a few questions, and about sixty seconds later they can read the entire book before they pay a cent. No deposit. No wait to see if it is any good. The full thing, up front.
I built that end to end. The design, the copy, the flows, the front-end, the way the agents hand work to each other and the way a human checks it before anything reaches a customer. There was no team I was handing tickets to. There was an idea about how a magical thing should feel, and there was me, making it real one piece at a time until it worked.
If I had only known how to draw the screens, Story Genie would still be a deck. The interesting problems only showed up once it was running. The preview had to feel like a gift, not a sales tactic. The quality checks had to be invisible and reliable. None of that is a Figma problem. All of it is a building problem, and I could only solve it because I could build it. I wrote up how I built it if you want the longer version.
Leading and building are not enemies
People assume that if you are still building, you must not be leading. I lead design as my day job, in a multi-billion-dollar eCommerce org, through an AI-first transformation that changed how the whole team works. One of the agents we built took a workflow that used to eat around three hundred and twenty hours and brought it down to about one. That is not a tweak. That is a different cost structure for the entire function.
I could not have led that change if I had taken the old advice and walked away from making years ago. You cannot direct a transformation you do not understand at the level of the actual work. If I had only ever talked about AI, I would have shipped slogans. Because I build with these tools myself, I knew what was real, what was hype, and where the leverage actually was.
That is the part the old ladder got backward. Building did not pull me away from leadership. It is the thing that made my leadership worth listening to. I have written more about why leaders should keep building, because this is the question I get asked the most, and the assumption behind it is the thing I most want to break.
A leader who can still build makes better calls. You estimate effort honestly because you know what the work costs. You spot a bad plan early because you can feel where it will break. You earn trust from your team because you are not asking them to do things you could not do yourself. None of that survives if you let your hands go cold.
The gap is the widest it has ever been
Here is the uncomfortable part. The distance between designers who can build and designers who cannot is now the widest it has ever been, and it grows every month. Every new model, every better tool, every drop in the cost of turning an idea into a real artifact widens it further. This is not a gap that closes on its own. It opens.
And it is not a gap in raw talent. I know brilliant designers with better taste than mine who are stuck, because every idea they have still needs a translator and a queue before it becomes real. Their taste is trapped behind other people’s time. Meanwhile a designer with less polish but the ability to build is shipping, learning from what shipped, and shipping again. Speed compounds, and they pull away.
I wrote a whole separate piece on why the gap is not skill anymore, because it is the most important shift to understand right now. The thing separating the top from the middle is no longer who has the best eye. It is who can carry an idea all the way to something real without waiting on anyone else.
The danger is the same speed that makes you valuable
Here is the trap nobody warns you about. The thing that makes building powerful is the same thing that makes it dangerous. When you can go from idea to working thing in an hour, you can also go from a bad idea to a polished bad thing in an hour. Speed does not check itself. It will happily carry your worst instinct straight to production and make it look finished.
I have watched smart people fall into this. They learn to build, they feel the rush of seeing something real on the screen, and they start mistaking output for progress. They ship five versions of a thing before they have asked once whether the thing should exist. The work looks busy. The work looks impressive. But taste went quiet somewhere around version two, and now they are just sprinting in a direction nobody chose on purpose.
This is the part where the old fear was not totally wrong. The worry was always that designers who code would stop thinking like designers. That does not have to happen. But it can happen, and the way it happens is subtle. You get fast, and fast feels like skill, and you forget that the hard part was never making the thing. The hard part was knowing which thing was worth making.
So the discipline is simple to say and hard to live. Building serves the idea. The idea does not serve your urge to build. Every time you reach for the keyboard, the question underneath should be whether this brings the real thing closer or just adds motion. Motion is cheap now. That is exactly why it is so easy to confuse with work.
The best builders I know are ruthless editors of their own output. They make a lot, and then they throw most of it away without flinching. They treat their first working version as a question, not an answer. They look at the thing they just built and ask what it is missing, what it is faking, and what a person on the other end will actually feel when they use it. Then they cut. Then they build again. The taste leads, and the building follows, every single time.
When I built Story Genie, the speed was the whole point and also the whole risk. Agents can write and illustrate a full personalized book in the time it used to take to sketch one page. That meant I could ship junk at a scale I had never been able to before. So I set one rule above all the others. Nothing reaches a family until it clears the bar I would hold a human author and illustrator to. The agents move fast. The standard does not bend. That is the only reason the speed is worth having. Cheap building without a stubborn standard in charge is not an advantage. It is just a faster way to make something I would be embarrassed to put a child’s name on.
Taste tells you what to make. Building lets you prove it. The designers who can do both are about to be the only ones who matter.
What to do about it
If you design and you cannot build, start now. Not next quarter. Pick one small idea and make it real this week with the AI tools sitting right in front of you. It will be rough. The first version always is. The point is not polish. The point is crossing the line from a picture of a thing to the thing, because once you cross it the first time, you never want to go back.
If you lead a design team, stop treating building as beneath your senior people. Reward the ones who ship real artifacts, not the ones who present the prettiest plans. Make it normal for a designer to bring a working version to a review. The teams that do this will move at a speed the deck-makers cannot match, and the difference will not be subtle.
And if you are trying to figure out what this means for your org, that is a lot of what I do now through AI design consulting. The companies that win the next few years will be the ones whose designers can build, and the ones that cling to the old ladder are going to wonder why their best ideas keep arriving late. The advice to stop making was always a little wrong. AI just made it expensive.