Learn to Code. Then Don't Stop There.

Photo by Desola Lanre-Ologun on Unsplash

Every young person in Kigali has heard the same career sermon by now. Learn to code. The internet doesn't care where you were born. Get a remote job, earn in dollars, work from your bedroom. Build your skillset and the money follows. It sounds like a door swinging open, and for a generation watching wages stay flat while everything gets more expensive, it sounds like the way out.

I'm a software student. I've done the remote work. I'm not here to tell you the advice is a lie, because it isn't. But there's a part of the sermon nobody preaches, and it's the part that actually decides whether any of this works for you.

The part where you overperform for an underwhelming job

Here's what they leave out about getting hired. You don't just compete for a job, you overperform to get a worse one than you deserve.

You study hard, you build the projects, you sharpen the skills everyone told you to sharpen. Then you go looking and the job you're actually qualified for already has someone in it. So you take the one a step down, where you're overqualified, doing work that doesn't stretch you, getting paid like the work doesn't stretch you either. To even land that one, you have to outshine a stack of other people who did exactly what you did. You overperform on the way in, and the reward is a job that underwhelms.

It isn't only happening to me, and it isn't only happening in tech. The numbers say the same thing in plainer language. Rwanda has done well bringing youth unemployment down, but the deeper problem isn't a simple lack of jobs. It's a mismatch. Most of the jobs available don't line up with the skills young people actually have, and youth make up more than half of the country's underused labor. More than 250,000 young Rwandans walk into the job market every year, and the number of formal jobs waiting for them doesn't come close. So a lot of qualified people end up either underemployed or squeezed into the informal economy, doing work that sits below what they trained for.

That gap between what you can do and what you can get is the first thing the sermon skips.

Remote work was never the great equalizer

The second promise is the shiny one. Work remote, the internet flattens the world, a kid in Kigali and a kid in California compete on equal footing for the same dollar.

I've lived inside that promise, and the equalizer has a catch. Remote work is real and it can be good. But a lot of it runs on a simple arrangement: they pay you the local floor and hand you a global workload. You're earning what someone here earns, while carrying what someone abroad would refuse to carry for that money. The dollars look fine on paper until you divide them by the hours and the volume and realize the deal was built to get the most out of you for the least. The internet didn't flatten the world. It just found a cheaper place to put the work.

That's not a reason to refuse remote work. It pays, and sometimes it pays better than what's local. It's a reason to walk in with your eyes open, knowing the arrangement is tilted, so you can negotiate, set limits, and never mistake being overworked for being valued.

AI didn't kill the path, it moved the floor

Now the part everyone's anxious about, me included. I'm studying to write code at the exact moment machines got good at writing it. You'd think that would feel like training to be a coachman the year the first cars rolled into town.

The doom is real in the data, I won't pretend otherwise. Employment for software developers aged 22 to 25 has fallen nearly 20 percent since 2024, according to Stanford's AI Index, even as headcount for their older colleagues kept growing. The first rung of the ladder, the junior jobs where you used to learn by writing the boring code, got thinner. Companies worked out that a few senior people with AI tools could ship what used to need a couple of juniors too. The grunt work that taught my generation the craft is the exact work the machines now do in seconds.

And it isn't only coding. The same report found the same pattern in customer service, where hiring for young workers dropped while the older ones held on. I spent time doing remote customer service, so this one isn't abstract to me. The work I was doing is exactly the kind that AI came for first. I've watched this from inside the field everyone's only now worried about, which is part of why I don't think the answer is to run from it.

But here's where I part ways with the panic. AI helps, enormously, if you know how to use it. The same reports that show the junior jobs shrinking also show who's still getting hired: the people who can direct the AI, check its work, and catch what it gets wrong. One engineer put it well, the best developers now aren't the fastest typists, they're the ones who know when to distrust the machine. Refusing to touch AI is the actual red flag in a 2026 interview. "I write everything by hand" doesn't read as principled anymore. It reads as someone who'll be slow.

So learn to code. Genuinely, you have to, and not because every line will be yours to write. Some things you'll always need to build and personalize yourself, and you can't direct or correct a machine on work you don't understand. Learn the craft so you can command the tool instead of being replaced next to it. The floor moved. It didn't disappear. You just have to stand higher on it from day one.

The part that actually matters: don't put all your eggs in one basket

Here's the advice underneath all the other advice, and it's the same lesson that runs through everything on this site. Don't stake your whole life on one move.

Everyone's shouting the same things right now. Build your skillset. Learn this, learn that. Good, do it. But do it and then keep going. If you're in tech, don't be only in tech. Learn another field next to it, because the most hireable people aren't the ones who went straight from a degree into a bootcamp, they're the ones who carry tech plus something else, a domain, a trade, a second lens on the world. One skill is a single point of failure. Two is the beginning of a foundation.

And don't build your whole life on a single wage, because a single wage doesn't stretch far enough anymore. This isn't some new discovery. It's how the generation before us already lives, quietly running more than one source of income because one was never enough. So once you've put away a bit of money, put it to work. Buy a piece of property and rent it out if you can get there. Invest the surplus instead of letting it sit and shrink. Treat your income the way you'd treat a portfolio, spread across more than one thing, so that no single failure can take the whole house down. It's the same logic I went through in the piece on what actually builds wealth here, just pointed at how you earn instead of how you save.

That's the part nobody puts in the sermon. Learn to code, yes. Take the remote job, yes. Build the skillset, absolutely. Just don't stop there, and don't bet everything on any one of them. The young people who make it through this aren't the ones who followed the advice hardest. They're the ones who took all of it, then quietly built a second floor underneath, and a third, so that whatever gives way, they're still standing.