Know Your Medium

For years, design discourse went like this:
should designers code?

It was a consistent conference talk, and an argument for either side made for perfect Twitter rage-bait. It generated more heat than insight, because it was always the wrong question. It assumed that coding was a secondary skill. A nice to have.

But while we were arguing about it, the tools moved. Design tools started thinking in systems, logic, and architecture. AI tools started shipping functional interfaces from a sentence. Somewhere along the way, the question stopped making sense. The line didn't blur, it evaporated.

And so, the old question is now irrelevant.

Designers must understand their medium the way a ceramicist understands clay. The ones who ship things that hold together understand the substrate. They know the line between expensive and cheap. They know that border radius wasn't a hill to die on but the interaction model of a flow absolutely was. They have taste about where to spend complexity. That knowledge can only come from proximity to the material. Not tutorials.

The old separation of designer draws picture, engineer makes picture interactive, was never a clean handoff anyway. Every layer of abstraction between intent and implementation introduced drift. The most functional, high-impact teams I've been a part of were the ones where the distance was shortest. Where a designer could look at a component and know without asking what it cost to change. Where an engineer could look at a flow and rapidly implement changes by understanding user impact.

That's not coding. It's fluency.

Now, the thought leaders have arrived, as they always do, late and ready to preach the hot topic of the time. The narrative is that now, everybody is a builder. The walls between disciplines are coming down. The future belongs to the "full stack product manager" or whatever gets the most impressions now.

They're not wrong, but they are treating a structural shift like a lifestyle choice.

What is actually happening is role collapse. The boundaries between product management, engineering, and design aren't blurring out of some democratising impulse. They're collapsing because the tools no longer enforce them. A designer who can prompt their way to a working prototype is making engineering decisions whether they call it that or not. A designer or engineer who understands the user and the business is making product decisions whether anyone gave them that title or not. The boundaries were always leaky. Now they're just gone.

But here's what I keep coming back to: this was always the trajectory for anybody who took product design seriously.

The role collapse isn't new. It's just louder. It's happening to everybody at once. Every good product designer I know eventually drifted into the liminal space between design, engineering, and product strategy. Not because they wanted to, but because good work demands cohesive inputs. You can't design a system without understanding how it works. You can't make meaningful decisions without understanding business priorities. You can't ship anything good without having opinions on implementation.

I spent years on that drift. Learning languages, frameworks, methodologies. Developing an intuition for what's possible and what is expensive. That knowledge was hard-won, and often hard to communicate to people who think design is pixel-work. But it's what made me useful in a room of engineers.

What's different now is that a designer with Cursor and good taste can produce something that would've taken significant effort five years ago. The decade it took to build that intuition is compressing into a year. That's not a complaint. That's the medium changing.

This job didn't exist 60 years ago. It won't exist in 60 years time.

The dissonance isn't really about this compression. It's about watching an entire wave of thought leadership describe a revolution, when for many of us it's just an acceleration of what we were already doing. We were already working in the medium.

So what actually matters now that the new tools are here?

Medium knowledge still matters, possibly more than ever. While tools make it easy to produce anything, producing the right thing is still as hard as it ever was. Taste and learning don't come from a tool. Judgement doesn't come from a prompt. Understanding the constraints and possibilities of your medium (the real ones, not the ones oneshotted in a demo) is still the work.

Role collapse is real, and it's not going away. But it's not a catastrophe. It's a recognition that roles were always somewhat artificial—a way to turn creating new things into a production line.

Indie hackers always existed, exist now, and will always exist. The titles we invented won't.

The question isn't whether designers should code, or whether PMs should be builders. The question is whether you understand the material well enough to make good decisions.

The ceramicist doesn't ask whether they should understand the kiln. They just understand it, because the alternative is broken pots and wasted time.

Know your medium.