For most of history, both sides of human input/output were expensive. Getting information into a person was slow — apprenticeships, libraries, years of study. Getting useful output back out was slower — craftsmanship, writing, analysis, all bottlenecked by the speed of hands and the limits of a single mind.
The internet blew open the input side. Suddenly, anyone could access almost anything ever written, recorded, or photographed. Information became nearly free. But the output side held. You still needed a human to synthesize, to build, to create. Knowing how to make something was cheap. Actually making it was still expensive.
AI is collapsing the output side.
When Output Gets Cheap
A skilled developer used to be valuable because the distance between "I need an app" and a working app was enormous. A designer was valuable because the distance between "I need a brand" and a coherent visual identity required years of trained judgment. A writer was valuable because clear thinking rendered into clear prose was genuinely rare.
AI compresses all of those distances. Not to zero — not yet, and maybe not ever — but enough that the economic floor drops out. When a single person with the right tools can produce what used to require a team, the supply of output explodes. And when supply explodes, price collapses. That's not a prediction. That's a law.
This is the knock-on effect nobody wants to sit with: the market value of human output — the thing most people sell for a living — is in structural decline.
The Knowledge Cliff
Knowledge falls the hardest. Knowing things used to be a career. Doctors, lawyers, analysts, consultants — entire professions built on the asymmetry between what they knew and what you didn't.
That asymmetry is evaporating. When anyone can query the sum of human medical literature in plain English and get a synthesized answer in four seconds, "I went to school for this" stops being a moat. Knowledge becomes a commodity. Commodities get priced accordingly.
What Goes Up
Supply and demand still work, even when everything else is shifting. So what becomes scarce when output is cheap and knowledge is free?
Good ideas.
Not "ideas" in the startup-pitch sense. Not shower thoughts or napkin sketches. The real thing: a clear vision of something that should exist but doesn't, combined with the taste to know what it should feel like and the judgment to know what to leave out.
Ideas have always mattered, but they used to be subordinate to execution. "Ideas are worthless, execution is everything" was the mantra of an era where execution was the bottleneck. When execution gets 10x cheaper, the bottleneck moves. It moves to the person who knows what to execute and why.
Taste. Judgment. Vision. The ability to look at a field of infinite options and choose the right one. These are the new scarce resources, and they don't come from training data. They come from lived experience, from giving a damn, from the kind of pattern recognition that only develops when you've built things, broken things, and sat with the consequences.
The New Equation
Human input is flooded. Human output is getting cheaper by the quarter. Knowledge is approaching zero marginal cost. But the capacity to want the right thing — to have a vision that's worth executing — is as rare as it ever was.
Maybe rarer. Because we spent the last twenty years optimizing for knowledge acquisition and output speed, and almost no time developing the muscles that matter next: conviction, taste, and the willingness to be wrong about something that hasn't been built yet.
The I/O has changed. The question is whether we noticed what got left out of the upgrade.
Wise Mountain builds tools for people who know what they want to make. See the work.