building is cheap. the trap i'm climbing out of.

in the AI era, every artifact i ship is an experiment. the real job is shipping with distribution in mind — not functionality first. i kept defaulting to functionality first. it's a trap.

Here's the pattern I kept falling into: I'd have an idea, I'd estimate it would take two weeks to build, and I'd start building. Three weeks later I'd have something that worked. Then I'd start thinking about how to get it in front of people.

The problem isn't the timeline. The problem is the order.

When building was expensive, it made sense to fully spec the thing before shipping it. The investment in distribution came after because the investment in the build was already sunk. But when building is cheap — when a working prototype takes days, not months — the expensive part flips. The expensive part is now distribution. Getting people to care. Getting them to show up. Getting them to come back.

So the mental model has to flip too. Distribution-first, then build. Or at minimum: distribution in parallel, not sequential.

What I'm trying now: before I write the first line of code, I write the post-launch tweet. If I can't write it in a way that would make me want to click, I don't build the thing yet. It's a cheap filter and it's already killed two projects that deserved to die before they consumed three weeks of my life.

The trap is that building feels like progress. It is progress. It's just not always the right progress.