AI Is the New Entry Level

I recently read about the Dreyfus model - how people move from novice to expert. It got me thinking about what happens to that path now that AI is everywhere.

At the beginning, the path is simple but slow. You follow instructions. You experiment. You make mistakes. And through those mistakes, you start to feel the problem from the inside - not just understand it, but recognize it. Over time, you stop needing the instructions. You develop intuition. You can look at a problem and know, without checking anything, where it's likely to break.

That's what the Dreyfus model describes. Not just skill levels - but how understanding moves from the outside in.

I can relate to this. In network engineering, when a problem comes up, I don't follow the checklist. I go straight to where I feel the problem is. I can't always explain why. I just know.

Part of learning as a novice is doing things by hand. Finding the error. Spending three hours trying to understand why something isn't working. And when you finally get it - something clicks. A gap closes. You realize why the theory didn't make sense when you first read it. You needed to feel the problem before the words meant anything.

AI makes that click harder to reach. It removes the friction. But without friction, the gap in understanding never closes. You get the answer in seconds - what's broken, why it's broken, how to fix it. Which is useful. But you move on without the click.

Everyone is talking about AI replacing junior engineers and entry-level developers. But if AI handles that layer - how does someone actually become an expert? The early stages aren't just inefficiency. They're the training. You learn by doing things wrong first. By spending time inside the problem.

AI may change the entry point. It can handle the tasks a novice once did. But it cannot replace the foundational knowledge that sits underneath - the part that tells you when something is wrong, and where to look.

That foundation is still the real entry level.