Most apps are built for millions. What if you could build one just for yourself? How AI is enabling a new era of personal software.
Published
Read time5 min
TopicProduct, Technology
TypeEssay
SeriesThoughts
I use about 10% of every app I pay for
You know the feeling. You sign up to do one specific thing (track a habit, forward a message, rename a batch of files) and the app does that thing on page three, buried under fourteen features you'll never touch. The pricing page has five tiers, and the one feature you actually want sits in the $19-a-month plan next to a hundred you don't.
So you pay for the whole thing, or you go without. For years, that was just how software worked. AI is quietly ending it.
Apps aren't bloated to annoy you. They have no choice.
Most "feature bloat" complaints miss the real cause. Apps aren't bloated because someone wanted to irritate you. They're bloated because they have no choice.
A company can't build software for one person; the math doesn't work. So it builds for the average of a million people, then bolts on feature after feature to catch as many of them as it can. Your specific, oddly-shaped problem, the one unique to your life, was never going to make the roadmap. You were always a rounding error. For most of computing history, "there's an app for that" really meant "there's an app for the most common version of that, and you'll bend yourself to fit it."
The day I just wanted an SMS forwarder
Here's the example that made this click for me. I wanted to forward certain text messages. That was the whole ask.
Not to email. Not to WhatsApp. No dashboard, no account, no subscription, no "Pro" tier. Just certain texts, forwarded simply, the way I had it in my head. Every app I found did too much, each with its own opinion about how forwarding should work, and none of those opinions were mine. The one feature I needed sat trapped inside three bloated apps, each asking a monthly fee for ninety-nine things I'd never open. A few years ago, that's where the story ends: you shrug, pick the least annoying option, and move on.
Now I can describe exactly what I want in plain English and have a working little app, mine, doing only my thing, in about a day. No real coding background. It does one job, my way, and costs me close to nothing.
That used to mean hiring a developer or learning to code. Now it's just a conversation.
You stop renting a mansion to use one room
And the cost barely registers. Most of the building blocks (APIs, integrations, messaging tools) are already sold in tiers, and the expensive tiers exist for businesses sending millions of messages. You're one person with one small need; you'll use a sliver of any free allowance.
So the whole equation flips:
**The old way:** pay for an entire app (and its whole pricing tier) to use one feature.
**The new way:** build the one feature, run it on free or near-free infrastructure, and own it.
You stop renting a mansion just to use one room.
This already has a name: microapps
When an idea feels this clean, I get suspicious of it. So I went looking to see whether real people are doing this or whether it's just a nice thought. They're doing it, and it already has names.
People call these tiny, single-purpose tools microapps: small, hyper-specific software built for one person's one problem. The habit of building them by describing what you want has picked up the (slightly silly) name vibe coding. And the deeper idea, software built for an audience of one, gets called software for one. A student tired of arguing with friends over where to eat built her own decision app in about a week, with no programming background. Developers are quietly building personal tools to kill the boring parts of their week, not to sell, just to use. It's already happening.
The real shift is who the code is for
We've had "AI writes code" headlines for a while; that's not the interesting part. The interesting part is who the code is for.
The threshold for "worth building" just collapsed. Problems that were always too small, too personal, or too weird for any company to bother with, the long tail of everyone's daily friction, are suddenly solvable in an afternoon.
Build it, use it, maybe throw it away
Some microapps will stick around for years. Others will be disposable, built for one trip or one project, then deleted. That's fine; we don't keep every grocery list either.
And more and more of them will be deeply personal: the forwarder that works exactly your way, the tracker that counts exactly what you care about, the tiny tool that does the one thing the big app refused to do without an upsell.
The honest caveats the hype skips
I'd be doing you a disservice if I pretended this is frictionless. When you build your own software, you also own its quirks. There's no support team; if it breaks at 11pm, that's you.
Anything touching your private messages, your accounts, or your money deserves real caution. Convenience is never a reason to be careless with personal data. And plenty of things still belong to polished, maintained, properly secured apps from real teams. Your bank should not be a weekend project. So the future isn't "everyone builds everything." You'll keep buying software when buying makes sense, and build it yourself when the need is small, specific, and unapologetically yours.
Software finally adapts to you
For decades, we adapted to software: learning its menus, paying for its tiers, tolerating ninety extra features to reach the one we came for.
Microapps reverse that. Now the software adapts to you, because you can make it. You don't need the whole app. You just need your one feature, built your way, for almost nothing. And increasingly, you can just go build it.
