Why I Build

There's a comfortable version of being technical where you read papers, understand architectures, and form opinions about trade-offs. You can go surprisingly far this way.

But there's a gap between understanding and building that no amount of reading closes.

When you build something, you discover the problems that don't appear in architecture diagrams. The ORM that generates 47 queries for what should be one join. The distributed lock that works perfectly until clock skew makes it not work at all.

I build because building is the highest-bandwidth way to learn. Not the most comfortable — reading is more comfortable. Not the most efficient — sometimes you build the wrong thing. But the most honest, because the system either works or it doesn't.