Let’s talk real. A decent software engineer follows a path that looks something like this:
- You start learning to think logically, building a foundation of concepts, figuring out how to actually code.
- Then you dive deep: pointers, memory, clean abstractions, OOP, design patterns — the good, the ugly, the confusing.
- Eventually, you move from academic code to real-world messes. You realize the goal isn’t elegant code, it’s solving the damn problem. But the way you code does matter — because ugly code bites back.
- Over time, you learn to swim through the chaos: tech, business, product, project politics. You learn when to write code, when to shut up, and when to say “this is f*cked.”
- Finally, you start laying down foundations. Others build on them. They make mistakes, same as you did. But now you’re steering the ship, not rowing at the back.
Now, the AI comes in.
It’s not here to replace you. It’s here to accelerate you.
If you’ve got critical thinking, AI is jet fuel. It is not perfect, sure, and it f*cks up things — but it does it in seconds, not days. You can tell when it’s wrong. You can fix it. And while fixing, you learn even more. Speedrun mode.
But here’s the kicker: AI assumes the human in the loop knows what they’re doing. That they’ve eaten enough sh*t to tell right from wrong.
And that’s not a junior. Not yet.
So now companies ask themselves:
Why hire someone slow, who makes mistakes, who needs hand-holding… when I can prompt Claude and get 80% of the result instantly?
The answer is simple: they won’t.
Juniors make you slower. AI makes you faster. Guess who wins.
And that’s why the junior dev market is toast.
Since there’s no turning back, maybe it’s time to rethink how we approach software careers — assume AI is here to stay, and focus on mastering the abstract stuff it still is not powerful at.