This question hit me hard the first time I really sat with it. We're the ones building the AI tools. We're training them on our own code. We're making them better at the thing we get paid to do.
Are we making ourselves obsolete?
It scared me. Of course it did.
But then I stepped back and looked at history.
This has happened before
When trains were built, entire industries around horse transport didn't just shrink — they vanished. Stagecoach drivers, blacksmiths, feed suppliers. Gone. When cars came, the train monopoly cracked. When planes came, long-distance rail declined.
Every single time, people thought it was the end. Every single time, it was growth.
Assembly lines replaced manual factory work. ATMs replaced bank tellers at the window. E-commerce reshaped retail. Streaming killed video stores. None of this was the apocalypse. It was the next step.
The pattern is always the same: the old way dies, but the people who adapt don't.
So are we shooting ourselves in the foot?
Kind of. But not in the way you think.
We're not eliminating our profession. We're eliminating the version of it that existed 3–4 years ago. The one where you spend hours writing boilerplate, Googling syntax, copy-pasting from Stack Overflow, and manually writing every single test.
That version is going away. And honestly? Good.
What remains is the hard part — the part that was always the real job but got buried under repetitive execution: deciding what to build, designing how it works, and making sure it actually solves the problem.
It's not about fear — it's about surfing
The instinct is to panic. To hold onto what you know. To resist the change and hope everything stays the same.
But everything changes at massive speed now. What we did two years ago is already outdated. The tools, the workflows, the expectations — all shifting constantly. You can't freeze time.
The choice isn't between "keep your old job" and "lose everything." The choice is between standing still on the shore and learning to surf the wave.
Surfing means adapting. It means accepting that your role will look different next year. It means investing in the skills AI can't replace — critical thinking, system design, understanding what to build and why. The planning, the strategy, the taste.
I wrote about this shift in my own work here. The change is real, but it's not a loss. It's a trade-up.
What actually stays
AI can write code. It can write tests. It can generate components and refactor functions. But it still can't:
- Decide what to build — understanding users, business context, tradeoffs
- Define quality — knowing when something feels right vs. just technically correct
- Own the outcome — being accountable for a system that works in production, not just in a demo
- Navigate ambiguity — the messy, human, "it depends" part of real projects
Those are the skills that matter more now, not less. The developers who invest in thinking clearly will outpace the ones who just learned to type faster.
The real risk
The real risk isn't AI replacing developers. It's developers who refuse to evolve being left behind by developers who did.
It happened with every technology shift in history. The blacksmith who learned to work with engines survived. The one who kept shoeing horses didn't. Neither was wrong about their skill — but one was wrong about the direction.
We're not shooting ourselves. We're growing. And growth always feels uncomfortable while it's happening.
Get on the board. The wave is already here.