Innis

Big companies go where the big markets are. They have to. The economics require it. But there are a thousand small markets they never touch, specific problems in specific industries that generic software handles badly, customers who are underserved because serving them correctly would cost more than they're worth to a company that needs scale. When the cost of building something drops far enough, the math changes on what's worth building. The one-person shop serving the market too small for anyone else to bother with is not a consolation prize. It is where the opportunity is.

Before, you had to learn to code or hire someone who could. You had to spend money on infrastructure before you knew if anyone wanted the thing. You had to make a bet with years of your life on an idea that might be wrong. Most people made a different calculation and took the job and stayed in it. The AI changes the number. Now the cost of finding out if you're right is measured in weeks instead of years, in evenings instead of savings accounts. When the cost of being wrong drops that far, the rational thing changes.

The man who fears the machine fears the wrong thing. He thinks his value was in the execution, in the years spent learning to write the code or draw the plans or compose the brief. But that was never the value. That was the cost of getting to the value. The value was always the seeing. What you noticed that others walked past. What you understood about a particular customer in a particular place that no textbook could have told you. The machine reduces the cost of execution to nearly nothing. What it cannot do is notice what you notice.

The machine cannot walk through your neighbourhood at seven in the morning and notice what's missing. It cannot sit in the back of a meeting and watch the thing everyone complains about and no one fixes. It cannot feel the specific frustration of a specific customer who cannot quite articulate what she needs but would recognise it immediately if you built it. This is your knowledge. Local, particular, yours. Hayek spent his career trying to explain why central planners always failed and the reason was simple: they didn't know what you know, and there was no mechanism by which they could find out.

Outside a city that was already lost, the enemy at the walls and everyone inside knowing it, and Jeremiah walking out to look at a plot of ground that would belong to foreign conquerors within the week. He paid the silver anyway. He signed the deed anyway. He sealed it in clay and gave it to Baruch for keeping and went back to the court of the guard and sat down. He understood that what a thing is worth now and what it will be worth across the long years are two different questions, and most men only ever think to ask the first one.

Creator to reader, value returned for value given, silver crossing between two hands in the morning light at Anathoth three thousand years ago and a satoshi moving across a channel tonight. Nobody takes a cut. Nobody decides whether the exchange is permitted. You made something. Someone found it worth something. The worth came back to you. Build enough of those and what you have is not a following but a set of actual relationships, and when the platform folds, the relationships don't fold with it because they never lived there in the first place.

First they need you, so they give you what you want. Then someone else needs them more and they give that someone what they want and you get what's left. Then another someone arrives. This is not betrayal in any useful sense of the word. It is just what happens when a business has to choose between the people who use it and the people who own it. The people who own it win. They always win. This should not surprise you. You built your house on their ground and you should have known whose ground it was.

Email is fifty years old and no one owns it and there is no executive who can decide one morning that the terms have changed. It runs the way weather runs, the way stone sits in a riverbed. A platform is a business with obligations to people who want their money back, and sooner or later those obligations change what the platform will do to the people it said it served. Build on the stone when you can. Borrow the scaffolding when you must. The scaffolding always comes down.

Silver passed between two men who could see each other's faces in the morning light at Anathoth and that was the whole of the transaction. No platform taking its portion. No algorithm deciding visibility. Deed pressed into clay. Witnesses standing by. The simplicity wasn't a primitive thing. It was a complete thing, with nothing borrowed and nothing owed to anyone who hadn't been there.

Platforms arrived and told creators they were partners and for a while it was true, and then it wasn't, and then the platform owned the audience and the creator was someone the platform pointed the audience toward when it served the platform's purposes. What had looked like a stage was something else entirely.

They built it so you would come back. The reward timed out just long enough that you'd need to check again, and in exchange for this arrangement you gave them your attention and your work and your connections and your years, and in the end they owned everything you'd built and you owned the memory of having built it.

The patient builder is their real problem, the one they can't solve. He doesn't need the number to go up today. He can wait, and a man who can wait is a man they have very little leverage over, and they know it.