I built a Scala tutorial in 2013. It was my baby—a full interactive course that ran in the browser. I poured weekends and late nights into it. Then I got a job, life got loud, and the repo went dark. For 11 years, it sat there, a digital ghost.
You know the feeling. That side project you started with fire, then buried under a pile of ‘I’ll get back to it someday.’ That someday never comes—until AI shows up at your door.
Last week, I dusted off the old code. Scala 3 had shipped, the original lessons were outdated, and the tech stack was a museum piece. But I had a new tool: Claude. AI doesn’t just write new code—it’s the digital archaeologist for your abandoned dreams. In one afternoon, I migrated the entire project to Scala 3, modernized the build system, and deployed fresh interactive lessons. The old HN post from 2013? I updated it with a link to the new version. The community woke up.
This isn’t a story about automation. It’s about resurrection. The hardest part of reviving a side project isn’t the code. It’s the courage to open the old files, face the mess, and remember why you started. The hardest part of reviving a side project isn’t the code. It’s the courage to open the old files. Claude didn’t just write the updates—it held my hand through the first commit. It transformed friction into flow.
Every developer has a graveyard of half-finished projects. A blog that never launched. A game engine that burned out. A tutorial series that stopped at chapter 3. We tell ourselves we’ll come back, but the activation energy is too high. The language evolved. The framework died. The syntax changed. It’s easier to start fresh than to revive the dead.
But AI changes the math. Your old code isn’t dead—it’s just waiting for the right tool to translate it into the present. I didn’t need to re-learn Scala 3 from scratch. I asked Claude: ‘Here’s my Scala 2 code. Rewrite it in Scala 3.’ It did. Then I added: ‘Now make the interactive lessons work with modern browsers.’ Done. The project that was a tombstone became a garden again.
This is the real magic of AI: not creating new things, but redeeming the old. The emotion hits differently. It’s not just productivity—it’s nostalgia mixed with empowerment. You look at your old work and think, ‘I’m not a failure for leaving it. I’m a hero for coming back.’
Take a side: I’m calling this brilliant. Not because AI is flawless, but because it’s the first tool that genuinely cares about your past. It doesn’t judge your old code. It doesn’t ask why you stopped. It just helps you finish. AI is the ultimate collaborator for your unfinished business.
Your old side project is waiting. It’s not gone; it’s just one prompt away. Open the repo. Let Claude in. And watch the ghost come back to life.
FAQ
Q: This sounds like a one-off success. Can AI really revive any old project?
A: Not every project, but any project with a clear goal and reasonable codebase. The key is that AI excels at migration and modernization tasks—updating syntax, fixing deprecated APIs, adjusting build files. If your old code still compiles conceptually, AI can bridge the gap faster than you can re-learn the old ecosystem.
Q: What's the practical takeaway for someone with a graveyard of side projects?
A: Pick the one that still makes you feel something—nostalgia, pride, or even guilt. Use a tool like Claude or GPT-4 to do a 'code audit' first: ask it to read your old project and suggest the biggest changes needed. Then commit to one session. The activation energy drops dramatically when you have a co-pilot that doesn't judge your past self.
Q: Isn't this just a fancy way to rewrite old code? Why not just write a new project?
A: Because the emotional value is in the resurrection. Starting fresh is easy. Reviving something you abandoned is a mirror to your own growth. You see the old mistakes, the naive architecture, the janky comments—and you get to improve them with your current skills. AI accelerates that reflection, turning a guilt trip into a celebration of how far you've come.