Somewhere on your GitHub account is a repository you were convinced would change everything.

It has three commits. The README still says "Work in progress."

Maybe it was a new framework. Maybe it was the startup that would finally replace your day job. Maybe it was just a side project.

You remember exactly how excited you were. You probably don't remember why you stopped.

If you've been writing software long enough, you don't just accumulate experience. You accumulate abandoned ideas: a graveyard of projects that never became what you imagined.

And that's perfectly okay.

We Judge Old Projects by the Wrong Metric

It's tempting to classify every unfinished repository as a failure. The app that never launched. The library that never got its first user.

But most projects don't fail because the code was bad. They fail because life happened. A new job, a child, changing priorities, running out of weekends, or simply realizing the problem wasn't worth solving after all.

None of that is an engineering failure. It's just reality.

Every Project Leaves Something Behind

A few months ago I built a script that counts how many of my GitHub repos haven't had a commit in X months. A meta-tool for this exact blog post, basically. The number it spat out was depressingly high. It didn't stop me from creating another repo that same week.

I can barely remember all the projects I've started. But I still recognize pieces of them in the software I write today: a caching strategy first tested in some forgotten repo, a deployment workflow that started life as a throwaway experiment, a UI pattern that survived three rewrites before finally landing in the right product.

Even the bad ideas left something behind. Sometimes knowledge. Sometimes confidence. Sometimes just the certainty that I'd never solve a problem that way again.

Projects disappear. Lessons don't.

The Graveyard Isn't Empty Because You Failed

It's full because you built.

Every engineer has ideas. Only some become code. Only a fraction of those survive. That's not depressing, that's engineering.

The alternative is worse: a perfectly clean GitHub profile. No experiments, no risks, nothing that ever seemed worth trying.

I'd rather have a graveyard than a museum.

Then AI Changed the Math

For years, the biggest cost of an idea was building it. You'd have an idea on a Tuesday and think: "this will probably take three months." Most ideas died right there, not because they were bad, but because building them was expensive.

That's no longer true. Today I can build a convincing MVP in a weekend, sometimes in an evening. Not production-ready, not something I'd deploy to customers, but good enough to answer the only question that matters: does this idea deserve more of my time?

That's a profound shift.

AI Didn't Make Great Ideas More Common

It made bad ideas cheaper to discover.

For every project that deserves six months of work, there are dozens that deserve exactly one afternoon. Before AI, I couldn't afford to find out. Now I can.

My graveyard is growing faster than ever, and I'm genuinely happy about that. Those projects didn't consume months of my life. They consumed a Saturday, answered a question, and quietly made room for the next experiment.

The Economics of Ideas Have Changed

The bottleneck used to be writing software. For many engineers, that's now the easy part. The hard part is deciding what deserves to exist.

AI can generate code. It can't tell you whether anyone actually needs what you're building, whether the problem is meaningful, or whether it's worth another week of your life.

As building gets cheaper, judgment gets more expensive.

Build More Things That Die

This sounds backwards, but we should normalize abandoning projects earlier. Not every idea deserves persistence. Some deserve a prototype, some deserve exactly two evenings before you confidently say: "this isn't worth pursuing."

That's not failure. That's progress.

Don't Be Ashamed of Your Graveyard

Open your old repositories. You'll cringe: terrible architecture, questionable naming, half-finished features, ideas that never had a chance.

But you'll also find the engineer you used to be. Every abandoned project taught you something your successful projects quietly relied on later.

The code may be dead. The experience isn't.

Some projects become businesses. Some become blog posts. Some become the foundation for the next idea that's actually worth finishing.

Keep building. Keep experimenting. And if your graveyard keeps growing, good. It means you're still digging for the idea that's worth the effort.