~/thoughts $ less ai-ate-the-juniors.md
AI ate the juniors (and I'm only half sorry)
2026-07-12
When I came into IT, the people who understood computers were treated like wizards. Not because we demanded it — because the knowledge was genuinely rare, and you earned it by reading, breaking things and sleeping next to a server. That era is gone, and I've made peace with it.
What replaced it was stranger. For about a decade, IT became a gold rush: people arrived not because they were curious but because the salary table went viral. Serious companies hired juniors who — and I have witnessed this with my own eyes — struggled with "Hello world" and couldn't start the stack they were allegedly developing on. An absurd theatre. It's like handing your car to a person whose prior mechanical experience is herding cows in a field. The companies didn't mind: everybody needs inexpensive hands, and value-per-hand was somebody else's KPI.
Then the models showed up
And now AI is quietly eating exactly that layer. From what I hear, nobody is rushing to hire juniors anymore — the model writes the boilerplate faster, doesn't attend standups, and never asks to work from Bali. The people who added little are being replaced by a tool that adds a little more, at a fraction of the cost. I'd call it cruel if it weren't so precisely fair.
Am I fully glad? No — half. Because somewhere in that eaten layer are the real juniors, the curious kind, the ones who would have become excellent. They now have to prove their difference from a language model on day one. That's a harsh entry exam. It's also, if I'm honest, the same exam my generation took — just administered by economics instead of a modem connection.
Where that leaves someone like me
I'm not pretending to be immune. I'm getting older; the head is not what it was at 25, and I can feel myself loosening the grip. Someday a younger person with more fire will come to take my place, and that's the correct order of the universe. So I look around, scan other directions, and think about where the next years of my time are worth investing.
Because that's the real lesson under all of it — the wizard era ended, the gold rush ended, and AI will end too, in the sense that it will become ordinary plumbing like everything else. The only strategy that survives every era is the same one that worked in 2005: be the person who actually understands the system. All the way down.