Why Coase Is Too Small for What AI Is Doing

This is the second piece in a three-part line of logic:
Blockbuster → Coase → The Flood.
The first piece broke the status frame. It showed how people often confuse the collapse of an old recognition structure with the collapse of value itself. This piece moves to the next defense people reach for once that first illusion starts to fall away: ownership. The instinct is understandable. If AI is built from human expression, labor, and knowledge, then it seems natural to ask who owns the value and who gets to bargain for a share of it. But that frame, serious as it sounds, still assumes that the thing being argued over is cleanly separable, ownable, and containable in a way that becomes much harder to defend the moment you look closely at what human expression actually is.

Before you can understand why “the flood” is not the danger, you first have to understand why the ownership frame cannot hold the event.

That frame has real appeal. It sounds rational, serious, and morally grounded. It says something like this: AI systems are being built by absorbing human expression, labor, knowledge, style, and data at massive scale. If that is true, then the people whose contributions made those systems possible should not be left empty-handed. They should have rights. They should have leverage. They should be able to negotiate. They should be able to license what is theirs and receive a share of the wealth built from it.

That argument is not absurd. In fact, it is one of the most intelligent responses available from inside the current economic imagination. It is an attempt to preserve dignity through bargaining power. It says: do not reduce people to recipients of scraps after the fact. Recognize their contribution at the level of ownership before the machine consolidates all the value.

That instinct makes sense. It deserves to be taken seriously.

But taking it seriously is exactly what reveals its limit.

The problem is not that the ownership argument is evil. The problem is not even that it is directionally wrong. The problem is that it presupposes a kind of world that does not cleanly exist in the domain it is trying to govern. It assumes that the relevant units of value are separable, ownable, and bargainable in a way that becomes much harder to defend the moment you inspect what human expression actually is.

That is where Coase enters.

Coasean reasoning begins from a world of distinct parties, identifiable rights, and negotiable transfers. If two parties affect one another and the rights are clear, then in principle they can bargain toward some efficient outcome. The important detail here is not the technical theorem itself. The important detail is the kind of reality the theorem imagines. It imagines a world composed of relatively bounded actors interacting through relatively bounded claims.

That works better in some domains than others.

If smoke from one factory damages a neighboring property, the parties are relatively legible. The harm is localized. The rights can at least be described. The bargaining object is imperfect, but not unintelligible.

Now compare that to human language, culture, and intelligence.

What exactly is the unit being owned?

That question is where the ownership frame begins to wobble.

Take a simple sentence written by a single person. At first glance, it looks like an individual output. It appears to belong to the person who wrote it. And in one sense, of course it does. They wrote that sentence. It came through their mind, their choices, their phrasing, and their act of expression.

But what made that sentence possible?

The language was not invented by them. The grammar was not invented by them. The meanings of the words were inherited. The metaphors embedded in the culture were inherited. The forms of argument they drew upon were inherited. The education that shaped their mind was inherited. The technologies through which they encountered prior ideas were social products. The categories they used to organize thought were built across generations. Even the sense of self through which they speak was formed relationally.

So yes, the sentence is theirs in one immediate sense. But it is also a local expression of a much larger distributed field.

That distinction matters.

The ownership argument becomes unstable because it quietly treats local expression as though it were produced in isolation. It treats the visible endpoint of a process as though it were the whole source of the value. But in domains like language, art, science, and cognition, value does not arise from isolated units in the way a strong property frame wants it to.

It arises through emergence.

Emergence means that something real and valuable comes into being through the interaction of many parts without being reducible to any one part alone. A market is emergent. A culture is emergent. A language is emergent. Scientific knowledge is emergent. Meaning itself is emergent. No single person creates these things alone, yet every person participates in carrying and reshaping them.

That is closer to what human expression actually is.

An individual contribution is real. It matters. It can be brilliant, original, courageous, transformative. But it is still not best understood as the production of a sealed atom of value detached from the shared structures that made it possible. It is better understood as a local crystallization of a much wider process.

This is not mystical. It is obvious the moment you look carefully.

No scientist creates in a vacuum. Even the most groundbreaking discovery depends on inherited mathematics, accumulated experiments, shared terminology, prior institutions, and collective memory. No novelist writes outside language. No musician writes outside genre, rhythm, instruments, influence, and audience. No thinker thinks outside a world already saturated with history, symbols, constraints, and prior forms.

The individual matters. But the individual is not the entire source.

That means the ownership frame is trying to force a distributed phenomenon into a legal and economic model built for more bounded objects.

And that mismatch has consequences.

If you push the property logic too hard, you begin treating participation in shared reality as though it were a set of fenced micro-estates. You begin imagining that every sentence, pattern, style, or idea can be traced backward into a chain of private claims clean enough to support bargaining at scale. But the more seriously you try to do that, the more absurd the object becomes.

Who owns a style influenced by ten thousand others? Who owns a pattern of reasoning carried through centuries of language? Who owns the cultural background that made a phrase legible? Who owns the scaffolding that made intelligence expressible in the first place?

The answer cannot simply be “everyone,” because that does not rescue the bargaining model. It overwhelms it.

And it cannot simply be “the final speaker,” because that ignores the distributed substrate from which the value emerged.

This is the trap.

The ownership frame sounds strongest at the level of intuition and weakest at the level of ontology. Intuitively, people feel that something of theirs has been used, and they are often right. But when you ask what the thing is, where it begins, where it ends, and how it can be cleanly separated from the shared field that produced it, the apparent solidity starts to dissolve.

That does not mean all rights claims disappear. It means they become secondary instruments rather than primary explanations.

This is the distinction that matters most.

Governance is not ontology.

You can still decide, as a society, that certain uses of data require consent. You can still decide that creators deserve compensation. You can still draw legal lines around privacy, attribution, licensing, and access. Those are governance decisions. They may be wise, necessary, and just.

But drawing lines for governance does not prove that the underlying reality is fundamentally composed of the lines you drew.

That is the mistake.

The ownership frame keeps sliding from “we may need legal protections here” into “this whole domain is fundamentally made of separable private property.” Those are not the same claim. The first may be practical. The second is often false.

And this matters enormously for AI because AI is not simply consuming isolated products. It is operating in the terrain of pattern, relation, language, inference, and abstraction. It is moving through the same distributed field from which human cognition and expression already arise. That is why the old frame feels increasingly strained. It is trying to treat intelligence as though it were a warehouse of disconnected inventory.

It is not.

Intelligence, especially at the level of civilization, looks much more like a living field of distributed participation. People are not merely owners standing outside that field. They are nodes within it. They shape it, draw from it, carry it forward, and are shaped by it in return.

That is why over-privatizing the event may create a new kind of blindness.

In trying to prevent exploitation, we may accidentally redescribe all knowledge as enclosure. We may train people to see every contribution primarily as a fenceable extractive claim rather than as a participation in a larger structure of intelligence. We may deepen fragmentation precisely when the event itself is revealing how interconnected the substrate always was.

Again, this does not mean exploitation is acceptable. It does not mean corporations should be free to take everything. It does not mean creators should shrug and accept dispossession. It means the highest frame cannot be exhausted by bargaining.

That is the real critique.

Coase is not too small because bargaining is silly. Coase is too small because the world it presupposes is smaller than the phenomenon now unfolding. It presumes bounded actors negotiating over bounded units of value. But intelligence, language, and culture are not merely piles of bounded units. They are emergent formations generated through countless overlapping contributions across time.

The property-rights response to AI is therefore best understood as a local defense mechanism inside the transition, not as an explanation of the transition itself.

That is why it eventually runs out of runway.

It can tell a story about compensation. It can tell a story about leverage. It can tell a story about governance and fairness. Those stories matter. But it cannot fully tell a story about what human intelligence is, where value in language comes from, or why the boundaries between self, system, contribution, and environment are becoming harder to maintain.

And once that becomes visible, the next step stops sounding abstract and starts sounding necessary.

The deeper shock of AI is not simply that people may lose bargaining power. The deeper shock is that the event is exposing how much of human life was never cleanly atomized to begin with. It is exposing the shared substrate beneath the illusion of total separateness. It is showing that intelligence was always more distributed, more relational, and more field-like than the ownership model wanted to admit.

That realization can feel threatening at first. People hear it as erasure. They think it means the individual no longer matters.

But the opposite is closer to the truth.

The individual matters precisely as a real local expression of a larger whole, not as a sealed sovereign container floating outside it. Human dignity does not have to depend on the fantasy that all value originates in atomized isolation. In fact, tying dignity too tightly to that fantasy may be part of what is now breaking.

What AI is forcing into view is not just a new market problem. It is a mismatch between our inherited economic language and the deeper structure of the domain we are trying to describe.

That is why the ownership frame feels both compelling and insufficient. It is reaching for justice with tools shaped by a thinner picture of reality than the event will allow.

So the right conclusion is not that property, consent, and compensation are meaningless. The right conclusion is that they are downstream. They belong to the level of social response, not to the deepest level of what is happening.

And once you see that, “The Flood” no longer sounds like a leap.

It sounds like the next necessary sentence.

Because before you can understand why the flood is not the danger, you have to understand why the walls you wanted to build around the event were never built to hold a living field in the first place.

If Blockbuster was about confusing medium change with humiliation, this is about confusing distributed emergence with separable property.

And that confusion, too, is about to fail.

But once the ownership frame breaks, the argument has nowhere smaller left to hide. It can no longer rely on status, and it can no longer rely on the fantasy that the event can be fully contained inside separable property claims. At that point, the real question finally comes into view: what kind of event is this, actually?

That is the next step. And it is the deepest one yet.

Next: The Flood Is Not the Danger