Field Note 031 · On the AI era · 7 min

AI made content cheap.
Structure is the advantage.

The cost of a competent paragraph went to zero. Almost everyone noticed, and almost everyone drew the wrong conclusion from it.

The reasoning ran: production is cheaper, therefore produce more. It is a defensible inference and I watched a great many capable teams act on it. Output tripled across the businesses I was closest to. Very little else moved.

What actually happened is subtler and worse for anyone who built their advantage on effort. Volume stopped functioning as a signal. When producing forty considered posts a month required a team, doing it proved something — resourcing, seriousness, competence. It was expensive, and the expense was the message. Now it proves only that you have an account with somebody.

When production is free, the constraint moves to discernment.

Where I was wrong

The correction is the useful part of this note.

For roughly a year I told clients the answer was more distinctive content — that if generated writing was going to be generic, the response was to be sharper, stranger and more particular, at the same volume.

That was half right and it cost people money. Distinctiveness matters, but I had kept the volume assumption intact underneath it, and the volume assumption was the thing that had broken. The clients who improved most did not get more distinctive at scale. They produced considerably less, and each thing they produced had somewhere to sit.

That is the difference structure makes. A business with a genuine point of view has a test for what to make, which means it also has a test for what to skip — and the second test is now worth more than the first.

The comparison I keep returning to

Two businesses in the same category, same year.

A
Forty pieces a month
Generated, edited, scheduled, competent. No argument connecting any of it. Impressions rose. Enquiries did not. Nobody in the market could tell you what the company believed, because it had not been asked to believe anything.
B
Eight pieces a month
Each one advancing a single position held consistently for eighteen months. Lower reach, materially better enquiries. When someone in that category has the problem, one name arrives already attached to it.

The second business is not working harder. It is working from a structure, which is the only thing that tells you the thirty-two other pieces were not worth making.

The principle

What travels.

Generation is solved. Selection is not, and selection is where taste lives — which is inconvenient, because taste cannot be bought in a subscription and does not scale by headcount.

Build the structure that tells you what not to make. It is the only part of this that a competitor cannot acquire in an afternoon.

The letter this came from

These begin as The Eight.

Eight short things on brand, growth and taste, every week. Four minutes. The pieces worth expanding end up here as Field Notes.