Narrow Enough to Live On

A first principle from 1776 · AI and the division of labor

Narrow enough to live on

How specific your work can be is set by how many people can reach you. AI is about to widen that reach for one person at a time, and the narrowest specialists will be the ones who gain.

Village one neighbour in ten: be broad City one in a thousand: a trade The reachable world one in a million: one kind of clock

how narrow you can be customers you needpeople who can reach you

Talent does not appear in the equation. Only the bottom line moves, and it is the line AI is moving.

The clock your grandfather wound every Sunday has stopped. The man at the repair shop on the corner turns it over twice and says he can take a look.

He fixes watches, mostly: batteries, bands, a cracked crystal now and then. He will keep it for three weeks, and you both know he is guessing.

Somewhere there is a woman who has opened four hundred of exactly that movement. She knows which spring goes first and which pivot wears oval. She is not on your corner. That small fact has decided how work is organized for as long as there has been work.

The principle

Adam Smith put it in the title of a chapter in 1776: the division of labour is limited by the extent of the market. In the lone houses and small villages of the Scottish Highlands, he wrote, a farmer had to be butcher, baker and brewer for his own family. A porter could find work only in a great town. A village cannot keep a specialist busy, so it gets a generalist. A city can, so it gets the specialist.

Strip it to arithmetic. Say you need two hundred customers a year to make a living. If two thousand people can reach you, you need one in ten of them, so your work has to be useful to one person in ten. You must be broad. If two million people can reach you, you need one in ten thousand, and you can be the person who does one thing. How narrow you can be is the customers you need divided by the people who can reach you. Talent, ambition and passion are not in the equation. They do not change the bottom line.

The word that carries the weight is reach, and reach was never only distance. The railway, the mail-order catalogue and the internet each widened it, and each time the trades split into narrower ones. But for a person selling her own skill, the internet widened the wire more than the reach. A distant customer still had to find her, trust her, explain the problem, wait for a quote, book a time, pay, and come back with questions. If he spoke another language, all of it happened twice. That work around the work cost her almost as much per customer as the job itself. So most people who could have been narrow stayed local and broad. The wire reached the world. Her hours reached the county.

What AI changes

AI does not make anyone better at clocks. It changes the bottom line of the equation, from both ends at once.

On the buyer’s side, a delegate turns “it stopped” into a description precise enough to match: the maker, the movement, the symptom, the photographs. Then it looks for the person whose work fits that description, not the person with the most reviews nearby. On the seller’s side, her delegate handles the intake, the quote, the calendar, the invoice, the follow-up and the translation. The work around the work used to grow with every new customer. Now it shrinks to a sliver per customer. Her reach stops being the county her hours could cover. It becomes the people whose delegates can find her.

The common story is that AI replaces specialists. The principle points the other way for any skill a machine cannot yet copy: it makes room for narrower ones. The person it squeezes is the generalist in the middle, because the generalist was protected by one thing, which was being nearby. Nearness stops counting for any part of a job that can be described, shipped or done on a screen.

The generalist was protected by distance. Distance is the input that changed.

That cost is plain and I will not soften it. The capable person who does a bit of everything in a small town has had a steady life because the next specialist was a day’s drive away. For the parts of that work that can be done at a distance, the protection ends, and the principle offers no replacement. What remains local is the part that needs hands. The man on the corner does not disappear. More and more, he becomes the hands of someone far away.

You photograph the clock. Your delegate asks two questions about the sound it made before it stopped, then names a woman in Porto who restores one family of movements and nothing else.

Her delegate answers in English before lunch, with a quote and a padded box on its way. She has customers in thirty countries and none on her street. Her niche is her address.

Five predictions

  1. Job titles get longer.

    The way people describe their work narrows until it sounds like a joke: not “bookkeeper” but “bookkeeper for family vineyards with seasonal crews.” The narrower the description, the better a delegate can match it, so narrow descriptions will win the work.

  2. Being described exactly replaces being seen widely.

    For specialists, marketing shifts from reaching many people to stating precisely what you do and what you do not do. A clear page that a delegate can read will beat a large audience that a delegate ignores.

  3. The local tradesperson becomes the hands.

    In physical work, the diagnosis leaves town and the hands stay. Expect the nearby repairer, nurse or electrician to work more often on a screen under the direction of a distant specialist who has seen the problem hundreds of times.

  4. Career advice flips.

    “Keep your options open” was insurance against a small market. As the market stops being small, the better advice becomes “go narrower than feels safe.” Breadth will start to read as a lack of a clear answer to what you are for.

  5. Language stops being a border for skilled service.

    A specialist who speaks one language will serve clients in several. The first visible sign will be firms of one with most of their customers abroad, and owners who say so without surprise.

Where I stand

I teach my students one sentence about careers: let your first employer be your first customer. This principle sharpens it. The first customer used to have to be close, because closeness was what made the work around the work affordable. My advice now is to pick a problem so specific it embarrasses you a little, and let the delegates do the finding. I wrote about the other half of this in The Part That Cannot Be Split: the part of a job that has to happen in order is the part that keeps its price. The narrow specialist is usually the one holding it. And it is the same arithmetic behind firms of one: one person can stand behind a very small slice of the world once the world can reach her.

The claim ledger

Already true
Smith’s limit has held through the railway, the catalogue and the internet: wider reach was followed by narrower trades. Delegates that draft, schedule, translate and invoice for one person are already in use.
What has to happen
Buyers hand their searching to delegates that match on fit rather than fame. Trust between strangers gets cheap enough for a delegate to check: guarantees, records of past work, someone who answers if it goes wrong.
Where I am probably wrong
If delegates match on popularity, reach widens but flows to a few big names, and the narrow specialist is no better off than on a search page today. If by 2030 people earning a living from one narrow skill with mostly distant customers are not plainly more common, I was wrong.

Back at the corner

It is 4:42 p.m. The man is still holding your grandfather’s clock. He is a good repairman, and he is guessing. Somewhere a woman has opened four hundred of them. For most of history the distance between you and her was the end of the story.

It is about to become a detail.

Background

Author: John Rector

John Rector is a Charleston-based entrepreneur, author, and AI strategist. He co-founded E2open, the supply-chain software company acquired for $2.1 billion in 2025, and in 2026 opened Charleston AI, a 3,000-square-foot lab that helps people and organizations understand and use artificial intelligence. He is the creator of The Reality Equation — a lecture series, book, and curriculum exploring attention, prediction, and how reality is experienced — and the author of more than two dozen books. He writes and speaks widely on artificial intelligence, attention, and the future of human work.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.