One in four companies in Y Combinator's Winter 2025 batch shipped products built with 95% AI-generated code. We read that number twice. These are not weekend projects. They are venture-backed companies selected by the most competitive accelerator on earth, and a quarter of them barely touched their own codebase. The cost of building software — the thing that defined who could participate in the technology economy for fifty years — has collapsed to near zero.
That collapse changes everything downstream. And almost nobody is prepared for what comes next.
Who is building software now?
Everyone. Or close to it.
63% of vibe coding practitioners have zero programming background. People who could not write a function last year are shipping full-stack applications by describing what they want in natural language. The term "vibe coding" went from inside joke to industry practice in under twelve months.
Lovable, one of several platforms enabling this, sees 100,000 new projects created every day. One hundred thousand. Per day. We keep staring at that number. The production bottleneck has not merely loosened — it has effectively vanished for a wide category of software.
The sensation of watching this happen up close is vertigo. We have spent decades in a world where "can you build it?" was the qualifying question. That question no longer qualifies anything.
What happens when supply becomes infinite?
Almost all of those 100,000 daily projects die in obscurity. This is the part nobody talks about at the AI conferences.
When everyone can build, building confers no structural advantage. Scarcity of supply used to create value — if you could make the thing, you had an edge, because most people could not. Now supply is functionally infinite. The scarce resource has shifted entirely to the demand side. Can you find the people who need what you built? Can you explain why it matters? Can you reach them before they drown in the other 99,999 projects launched the same day?
Distribution is the moat, not production. We will say it plainly because the old mental model dies hard: the ability to build has been democratized. The ability to reach the right people, at the right time, with the right message, has not.
How big is the creator economy — and who is actually winning?
The direct creator economy has reached $178 billion, with projections to $1.35 trillion by 2035. There are 29.8 million solopreneurs in the United States alone, generating $1.7 trillion in combined revenue.
Both things are true at once. Individual value creation at scale is real and accelerating. And the distribution of outcomes within it is brutally uneven. A small percentage captures the vast majority of revenue. The long tail is not just long — it stretches past the horizon, populated by millions of builders who never reach critical mass.
The pattern holds across every platform and medium we have examined. Production tools get cheaper. Supply explodes. The median outcome for any individual producer drops. Winners are the ones who solved distribution. Not the ones who solved production.
This is not a prediction. It is already the landscape.
What did the anticivilization train people to do?
Exactly the wrong thing for this moment.
The zero-sum institutional structure most people grew up inside ran a training program with three steps:
Follow instructions. Do what the teacher says. Do what the boss says. Comply with the process. Collect credentials. Get the degree, the certification, the title. Wait for permission. Apply for the job. Apply for the grant. Wait to be selected.
Every step assumes the bottleneck is capability. You follow instructions to build capability. You collect credentials to prove it. You wait for gatekeepers who control access to production resources.
When AI hands everyone production capability overnight, this entire sequence becomes not just irrelevant but actively harmful. People trained to follow instructions are the least equipped to decide what to build. People trained to collect credentials discover that credentials do not solve distribution. People trained to wait for permission find that the gates are open, the field is infinite, and nobody is coming to tell them which direction to walk.
We say this without contempt for the people caught in the transition. The training was not their choice. But the mismatch between what they were trained for and what the moment demands is severe, and pretending otherwise does not help.
What is the question nobody learned to answer?
Not "can you build it?" That question is settled.
The question is "should you build it?" And underneath: Who is it for? Why do they need it? How will they find it? What changes in their life when they have it?
These are direction questions. They require self-knowledge — what are you specifically positioned to create? Market awareness — where is there genuine unmet need? Strategic clarity — what is the smallest thing you can build that reaches the right people? Distribution thinking — how does your audience currently discover solutions?
None of this was in the curriculum. The anticivilization produced excellent followers and competent producers. It did not produce people who can choose direction. Direction is now the only scarce resource.
Why do most solo builders hit a $50K wall?
Our research across solopreneur communities surfaces a pattern so consistent it feels structural: a $50,000 annual revenue wall. Below that line, you can sustain yourself on production alone — building things, selling services, freelancing. Above it, you need distribution infrastructure: audience, positioning, systems, strategic direction.
The builders stuck at $50K are often extraordinarily capable. They can build anything. What they cannot do is consistently reach and convert the people who need what they build. Production-rich. Direction-poor.
This is the inverted bottleneck at the individual level. And it explains how the creator economy can be simultaneously enormous ($1.7 trillion) and deeply frustrating for most of its participants.
Is this actually a positive-sum problem?
In the zero-sum frame, the inverted bottleneck looks like a competitive crisis. More supply, more competition, lower prices, race to the bottom. That reading is not wrong. It is incomplete.
In the positive-sum frame, the inverted bottleneck is an alignment opportunity. If the scarce resource is direction, then systems that help people find their direction create enormous value — not by competing for existing demand, but by activating latent demand that production constraints previously suppressed.
Think about it this way: there are problems in the world that nobody is solving, not because the solutions are too hard to build, but because the people who could solve them do not know the problems exist. The inverted bottleneck means the cost of solving any identified problem approaches zero. The entire value chain collapses to two activities: identifying the right problem and reaching the people who have it.
We do not yet know what the full infrastructure for direction looks like. We are building pieces of it and watching others build their own. But the structural logic is clear: the world does not need more production tools. It needs direction systems. And the people who build them — or who master their own direction — will define the next decade.
The bottleneck flipped. The question is whether you noticed.
This is Episode 7 of the Superpuzzle Developments series. Next week: "The dark flow" — what happens when capacity amplifies misdirection.