I used to build things because I had to.
Need a portfolio? Build something. Want to learn React? Build something. Trying to land your first job? Build. Something.
But then AI coding assistants showed up — and suddenly, anyone can build something. Your non-technical friend can prompt their way to a working to-do app. A designer can spin up a full-stack side project over a weekend. A product manager can ship a Chrome extension without writing a single line by hand.
So the question a lot of developers are quietly asking themselves now is: if AI can build it, why should I?
It's a fair question. And I think most answers out there get it wrong.
The Wrong Argument for Personal Projects
The usual case goes something like: "Personal projects show employers you're passionate. They prove you can build things outside of work."
That argument is getting weaker by the day.
Because if the bar is just having a project, that bar is now on the floor. GitHub is flooded with AI-generated CRUD apps, half-finished SaaS clones, and portfolio pieces that were built in an afternoon with Cursor and zero real understanding. Recruiters know this. Senior engineers know this.
A personal project by itself doesn't prove much anymore. So if you're building just to have something to show — you might want to rethink the strategy.
What AI Actually Changed
Here's what's actually different now:
AI lowered the cost of starting, but not the cost of thinking.
Spinning up a boilerplate? Instant. Writing repetitive CRUD logic? Done. Debugging a silly syntax error at 1am? AI's got it.
But figuring out what to build, why it's worth building, and how to make real decisions under constraints — that's still entirely on you.
AI is an incredible executor. It's a terrible strategist.
And here's the uncomfortable truth: a lot of developers have been hiding behind execution for years. Personal projects were a way to stay busy without doing the harder thing — which is developing taste, judgment, and genuine problem-solving ability.
AI just made that hiding place disappear.
The New Value of Personal Projects
Personal projects still matter — but the value has shifted.
1. They're now a thinking gym, not a proof of execution
When you build something yourself, even with AI helping you code, you're forced to make hundreds of small decisions. What's the data model? How do I handle edge cases? What do I cut for v1? What actually makes this useful?
Those decisions build intuition. And intuition is the one thing AI can't give you by reading your prompt.
The developer who has shipped ten side projects — even messy, unfinished ones — has a mental library of tradeoffs that no amount of AI output can replicate.
2. They reveal what you actually find interesting
This one is underrated.
When you're not building for a job or a client, you build what genuinely interests you. And over time, that pattern tells you something important about where you should be pointing your career.
If you keep coming back to developer tooling, maybe that's your niche. If your side projects all involve data and visualization, that's a signal. AI didn't change this — if anything, it amplified it, because now you can explore faster.
3. They're the best way to learn how to use AI well
Using AI to code is a skill. And like most skills, you get better at it by doing real things with stakes, not just toy exercises.
When you're building your own project, you hit real walls. The AI gives you something that almost works. You have to figure out why it doesn't. You learn to write better prompts, to know when to trust the output and when to be suspicious, to steer the model instead of just accepting what it gives you.
That meta-skill — knowing how to work with AI effectively — is increasingly what separates good developers from great ones. Personal projects are the best place to develop it.
4. Finishing still matters. A lot.
AI makes starting easier than ever. Which means the people who can actually finish things — who can push through the boring middle, make the hard cuts, and ship — stand out more, not less.
A completed personal project, even a simple one, signals something that a half-built AI-scaffolded repo doesn't: follow-through.
What Makes a Personal Project Worth Building Now
Given all this, here's a simple filter I'd suggest:
Build it if at least one of these is true:
- You're solving a problem you actually have (even a small one)
- You're exploring a technology or idea you're genuinely curious about
- You're trying to develop a specific skill you can't practice at work
- It's something you'd actually use, or that someone you know would use
Be honest if you're building it just because:
- You think it'll look good on a resume
- Someone on Twitter said this stack is hot right now
- You feel guilty for not having a side project
The first list produces projects that teach you things. The second list produces projects that drain you and end up abandoned in a private repo.
The Real Threat Isn't AI. It's Passivity.
The developers I see struggling in the age of AI aren't the ones who can't code fast enough. They're the ones who've become passive — waiting for AI to tell them what to do next, accepting the first output, never pushing back, never experimenting.
Personal projects, at their best, force you out of passivity. You have no ticket. No spec. No senior dev to ask. Just you, a problem, and a blank file.
That experience — uncomfortable as it is — is increasingly rare and increasingly valuable.
So, Are They Worth It?
Yes. But the point was never really to prove you can build things.
The point was always to become someone who thinks clearly, ships consistently, and knows what they're doing and why.
AI didn't change that. It just made it more obvious.
If this resonated, I'd love to hear what you're working on — or what side project you abandoned that you think about sometimes. Drop it in the comments.
Top comments (0)