A few months back I built myself a token accessibility checker. It sits in my design system workflow and speeds up accessibility checking across tokens, simulates how layers render across brands, and lets me test new palettes without committing them. Work that used to eat hours of my time and cause a fair few headaches.
It's ugly. Spacing's off in places. The UX hasn't been thought through beyond "does the button do the thing." If a client saw it, I'd want the floor to open up.
My reasoning at the time was simple: it's just for me. Function over form. Save the hours, move on. Who cares.
Except it's been eating at me since.
Why does that bother me?
That's the actual question, because "it's just for me" should have settled it. It didn't. I keep circling back to the thing, wanting to fix the spacing, wanting to give it real states, wanting to make it good, even though the only person who will ever open it is me.
If function-over-form was really a satisfying trade-off, I wouldn't still be thinking about it months later.
The client version of me wouldn't allow this
Here's the part that actually complicates things. On client work, I wouldn't ship this. Not because the client would necessarily notice the spacing, but because I wouldn't let it leave the building looking like that. Somewhere in me there's a line, and this tool is nowhere near it.
So what's the line actually measuring? For a long time I'd have said "someone else is watching, so the standard goes up." But I don't think that's it, or not all of it. Nobody is watching this tool, and it's still bothering me. Which means the standard I hold for craft isn't only a performance for an audience. Some of it is just mine.
I'm not the only one asking this
I read a piece on Microsoft's design blog recently, The naked designer, that's been sitting with me. The writer describes building a mental-health tracker for himself with AI in an afternoon: it worked, it genuinely helped, and it bothered him ever since, because finishing it was so effortless that he couldn't tell anymore whether the thing helped him or whether he just enjoyed the act of finishing it. He calls these private, bespoke, single-user tools "private fires," each one fitted perfectly to the person who lit it, warming no one else.
My situation is a bit different. I'm not unsure whether my checker helped, I know it did, it's saved me real hours, repeatedly, for months. That's not a feeling I have to interrogate. But the discomfort he's describing and the discomfort I'm sitting with are close cousins: something was made, it works, and there's still an itch that says this isn't finished, and it should be shared.
Private fires aren't new. Software's just new to them
The more I sit with it, the less I think this is an AI problem, or even a software problem. People doodle margins nobody will frame. They rig up a fix for a sticking door that would embarrass them if a joiner saw it. They organise their own kitchen in a system only they'll ever navigate, alter their own clothes with stitching only they'll ever inspect up close. None of that was ever built to be shared, and none of it needed to be to be worth doing.
Private fires aren't new. They're just new to software.
That's the bit I keep landing on. AI didn't invent the private fire, it just made software cheap enough to finally have one, the same way we've always had one for a doodle or a wonky shelf. What's actually new is doing this at scale, in a category of thing that used to have to be shared to exist at all, because building it was too expensive to do just for yourself.
The itch I can't place
So maybe the guilt is misdirected. Maybe an ugly tool that quietly saves me hours every week is exactly as legitimate as a bad drawer-organising system nobody else needs to understand.
But I still want to go back and make it properly good. Polish the states, fix the spacing, maybe even put it out into the world for other designers running their own systems.
Here's what I can't work out: is that itch a genuine signal that this thing deserves to exist properly, and I'm right to trust it? Or is it just the reflex of being a designer, the same instinct that makes me tidy a Figma file nobody's going to open? I don't know if that itch is universal, something everyone who makes things privately eventually feels, or whether it's a professional tic specific to people like us, who've spent years being told the work should always speak.
If you've built something just for yourself and felt that pull to go back and finish it properly, for no audience at all, I'd like to know: was it duty, was it habit, or was it something else entirely?