Notes from the Collapse
The Seventy Percent Problem
On the number that quietly decides whether a profession survives.
June 1, 2026 · 2 min read
A friend handed an AI tool a brief he gets paid forty thousand dollars to answer. It took the machine ten minutes. It took him three weeks. The output, he judged, was about seventy percent as good as his own.
He told me this as a complaint, but it isn't one. It's a measurement. And the measurement is the most important number in this whole transition, because seventy percent is precisely the figure that does the damage. Ninety-five percent would be a crisis everyone could see. Forty percent would be a curiosity nobody would pay for. Seventy percent is the number that hollows a profession out from the inside while letting everyone tell themselves the work is still safe.
Why seventy is the dangerous number
Most paid work does not require excellence. It requires sufficiency, delivered on time, at a price the buyer can stomach. The expert has spent fifteen years closing the distance between good and great — the last thirty percent — and has come to believe that distance is the job. To the buyer, much of the time, it is rounding error.
When a free tool clears the bar of sufficiency, the expert is no longer competing on quality. They are competing on the marginal value of the final thirty percent, against a price of zero and a turnaround of minutes. That is a brutal market to be standing in, and most people standing in it do not yet know they are.
Where the thirty percent still lives
The argument of the book is not that the thirty percent is worthless. It is that the thirty percent has moved. It no longer lives in execution — the drafting, the rendering, the first pass. It lives in knowing which problem is worth solving, in carrying the responsibility when the answer is wrong, in the taste that decides what to keep.
Those things survive. But they are a smaller share of the job than the people who hold them would like to believe, and they cannot be billed by the hour. The career that comes out the other side of this is built on the thirty percent that machines can't reach — which means it is built on a narrower foundation than the one it replaced.
That is the work ahead: not defending the seventy percent, which is already gone, but finding out how much you are actually worth once it is.