There's a specific kind of restlessness that only developers understand. A quiet, persistent itch. A mental background process that runs while you're eating, showering, or trying to sleep. Milan Jovanović called it "the urge to build something," and that's exactly what it feels like.

I've had that itch since I wrote my first line of code. And in 2026 — with AI tools that can scaffold an entire application in minutes — I've noticed something strange: the itch hasn't gone away. If anything, it's gotten stronger.

What side projects actually teach you

Your day job teaches you to be reliable. You work within existing systems, follow established patterns, and deliver what's expected. That's valuable, but it's not the full picture.

When you build something from scratch — when you own every decision from the database schema to the button colors — you learn differently. There's no senior developer to ask. No established patterns to follow. Every problem is yours to solve.

Through side projects, I've wrestled with authentication flows, payment integrations, deployment pipelines, and performance bottlenecks. I've built features nobody asked for and skipped features everyone needed. I've launched things that got zero users and a few things that found an audience. The outcome was almost never a successful business. But the person who finished those projects was not the same person who started them.

The AI twist

Here's the interesting part: AI tools have made side projects more valuable, not less.

In 2026, barriers that used to stop developers from building — "I don't know mobile development," "I've never done deployment," "I don't know that language" — have largely collapsed. You can prototype an idea in an evening that would have taken a week two years ago.

That acceleration isn't making side projects obsolete — it's making them accessible to more people.

The hidden ROI

There's a story that stuck with me. A developer spent seven months building an AI storytelling app. It got zero users. By every conventional measure, it failed. But during those seven months, she solved complex context management problems, built autonomous AI agent systems, and learned when to use a tool call versus raw generation. When she interviewed for an AI Engineer role, she walked the hiring manager through the architecture live. She was hired in under 30 minutes.

Her conclusion: "A project with zero users is not a failure if it gets you exactly where you need to go." That's the hidden ROI. The output isn't the product. The output is you — transformed by the act of building.

What I've learned

Start smaller than you think. The project that actually ships is the one that's narrow enough to finish. Scope is the silent killer of side projects.

Build for yourself first. The moment you start building for "users" you haven't found yet, you add stress without reward.

Don't wait for the brilliant idea. Starting gives you ideas. Waiting for the perfect idea is just another form of procrastination.

The transformation is the product. The skills, judgment, and confidence you gain are real. They transfer to everything else.

I keep building because the urge hasn't faded. And honestly? I hope it never does.