I can build almost anything now, which is the problem

A friend asked me a simple question this week: now that you can build an app in a weekend, what do you build?

I didn't have an answer, and the more I sat with it the worse it got.

Building used to be the wall. It isn't anymore. I can describe an app and watch it come together over a couple of nights. So can a lot of people. Walk through any maker community and you see the result: thousands of projects, most of them abandoned. The shelves are full. The real successes are a small minority, and the rest of us are mostly showing each other our demos.

So the question stops being "can I build it" and becomes "what is worth building." That one is much harder than it looks.

There are three textbook answers. Build something nobody has made. Build something that exists, but cheaper and simpler. Build something that exists, but for a narrow niche, solving one pain better than anyone. All three are true, and all three go blank the moment you actually ask "okay, so what." I have been staring at that blank for a while.

You can research your way toward it. Read the forums, read every review, scrape the stores, watch what real users complain about. I do that. But even with the data in front of you, the final call is still yours: is this worth building, can I actually build it even with the AI, do I want to, and do I know enough about it to tell when it is wrong. The data doesn't answer those.

When I look at the people who actually win, they share something I keep avoiding: they build the thing they understand and already need. They scratch their own itch first, then grow it out. It works because they know what they want, they have the domain knowledge, they know other people with the same problem, and the pain was real enough that they were going to make it anyway. They like the area. They were never guessing.

Then I look at myself, honestly, and it falls apart. My pains aren't painful enough to force me to act. The ones that are aren't general enough to matter to anyone else. Some are just work tasks I stop caring about after five o'clock. And the ones that might be real, I don't have the domain for. A guy mixing concrete needs to know the suppliers and prices in his area. I don't. If I built that, it wouldn't be my product, it would be outsourcing, and the people in the trade would do it better.

There is a quieter version of this, specific to how I work. I don't read the code my AI writes. I gate its decisions. I can only gate what I understand well enough to know when the answer is wrong. So the set of things I can build well is the set of things I already live inside. Same conclusion through a different door: the work I can verify is the work that is already mine. Right now I am short on those.

So for now I build small things that fit me, and I let them stay small. I don't fake a pain I don't have, and I don't copy a formula that worked for someone whose life isn't mine.

But I don't think I am stuck. I think I am early. Nobody is born with the thing they were meant to build. The people who have it got into a field, used the tools every day, and hit the same problem enough times that it became theirs. That I can do. Every project I ship, every room I get into, every subject I actually learn instead of skim, widens two things at once: the odds of running into a pain that is mine, and the depth to solve it when I do. Building stopped being the test. Becoming someone with something real to make is the test now, and unlike the blank question my friend asked me, that one I know how to work on. I keep going, and I get a little closer each time.