For twenty years I was told my scattered resume was a problem to be fixed.
Recruiters had a tell. They would scan the page, and somewhere around the third career change their eyebrows would do a small involuntary thing. Finance, then design, then a software venture, then a stretch building physical product, then back to something else entirely. The polite ones called it "eclectic." The honest ones asked, gently, whether I had trouble committing. I learned to apologize for the shape of my own life — to walk into rooms pre-defending the very thing that made me useful in them.
The advice was always the same. Pick a lane. Go deep. Become the person who knows one thing better than anyone. And it was good advice, for the world that was issuing it. In a world where each skill is expensive to acquire and slow to deploy, the specialist is the safe bet. You hire the cardiologist for the heart and the structural engineer for the beam, and you do not ask either of them to do the other's job. Depth was scarce, so depth was paid.
I want to suggest that this world is quietly ending, and that the people who were penalized by it are about to be repriced.
The bridge nobody photographs
Think about how a bridge actually gets built. There are people who understand soil — how it settles, where it lies, what it will and won't bear. There are people who understand steel — its tolerances, its fatigue, the temperature at which it stops behaving. These are deep specialists, and you want them deep. But soil does not speak steel. The ground engineer and the metallurgist do not, left alone, produce a bridge. Someone has to stand in the space between them, fluent enough in both to know when one is making a promise the other can't keep.
That person is rarely the most knowledgeable in the room about anything. Ask them to derive the steel's fatigue curve from scratch and they'll fumble. Their value isn't depth in a single material. It's that they can hold two materials in their head at once and see the joint — the place where the project actually succeeds or fails. We don't put their name on the plaque. We just couldn't have crossed the river without them.
That is what a generalist is for. Not a dabbler, not a jack-of-all-trades in the dismissive sense. A bridge engineer. Valuable precisely because they stand between two things that don't speak the same language.
What the tools just did
Here is the part the recruiters couldn't have known to weigh, because the ground moved underneath them this year.
Until very recently, the cost of any single skill was high. Becoming competent at writing, at coding, at financial modeling, at design, at legal drafting — each cost years. So the rational move was to amortize that cost over a lifetime in one lane. Breadth meant paying the acquisition cost many times and mastering none. The math punished the generalist, and the math was right.
The tools available now have started to collapse that cost. I can sit down with a language model and produce a competent first-draft contract, a working script, a clean financial schedule, a passable design brief — in an afternoon, in domains where I am nowhere near expert. I'm not pretending the output is a specialist's output. It isn't. But it's the floor that used to take a junior specialist months to reach, and it's available to anyone, instantly, at the cost of asking well.
Run that forward and notice what it does to the value of a narrow skill. When competence in any one domain gets cheap, competence in any one domain stops being scarce. And scarce is the only thing the market ever really pays for.
Where the value moves
Here's the thing the doom narrative misses. When the cost of a thing falls, value doesn't vanish. It moves to whatever the cheap thing is now bottlenecked on. Cheap steel didn't end engineering; it moved the hard part to design. Cheap computation didn't end thinking; it moved the hard part to deciding what was worth computing.
So when narrow skill gets cheap, what becomes the bottleneck?
Judgment about which skills to apply, in what order, toward a goal that no single discipline can see. The model will draft me a contract, but it won't tell me that the contract is the wrong instrument for this relationship. It will write the code, but it won't notice that the feature contradicts the business model. The tool answers the question you ask. It has no opinion about whether you're asking the right one. That opinion comes from standing between domains — from having enough finance to smell a bad deal, enough product to know what users actually do, enough engineering to know what's cheap to build and what only looks cheap.
The bottleneck moves to the joint. To the bridge engineer. To synthesis.
I spent decades collecting domains the way other people collect grievances, and feeling vaguely guilty about each one. It turns out I was accumulating exactly the thing that gets scarce when the tools make everything else common. Not depth in one material — the ability to hold several at once and see where they meet.
A caution, because breadth has a failure mode
Let me be honest about the trap, because I've fallen into it. Breadth without judgment is just noise. The generalist's real failure mode isn't shallowness — it's the seduction of the chase, the move to a new domain precisely when the current one gets tedious, which is exactly when the value is about to appear. I know that pull intimately. The difference between a bridge engineer and a tourist is that the engineer stays long enough in each domain to learn what it can and cannot bear. You don't need to be the best metallurgist alive. You do need to know real steel from a salesman's brochure. That floor of genuine competence is non-negotiable, and the tools don't supply it. They make the floor cheaper to reach. They don't let you skip it.
The repricing
For most of the industrial era, organizations were built as collections of specialists, and the org chart was a map of silos. That made sense when each silo was expensive to fill and slow to change. I suspect the next decade quietly rewards a different shape — fewer people, each holding more of the picture, with tools doing the narrow execution that used to require a department. Not because breadth is suddenly fashionable, but because the economics that punished it have inverted. When execution is cheap, the premium is on knowing what to execute and why, and that knowledge lives in the spaces between.
I don't say this as prophecy. I say it as someone watching the price of his own oddly-shaped career move, after twenty years of being asked to apologize for it.
If you've been told to pick a lane, I'd hold that advice up to the new light before obeying it. Go deep enough in your domains to tell the real thing from the brochure — that part still matters, maybe more than ever. But don't mistake the silo for the structure. The bridge was never made of soil or steel. It was made at the joint, by the person fluent enough in both to stand there. For a long time we didn't pay that person well. I think we're about to.