The Bifurcated Economy
The Coasean Boundary
Ronald Coase asked a question in 1937 that economics had been too busy with other matters to notice: if markets are efficient, why do firms exist at all? Every task should be coordinated through the market, every worker an independent contractor, the firm as an organizational form unnecessary. His answer was transaction costs. Using the market is not free. Discovering prices requires search. Negotiating contracts requires attention and legal skill. Specifying every contingency in a complex arrangement is prohibitively expensive. Monitoring specialized performance and enforcing agreements when someone fails to perform can cost more than the transaction is worth. When these costs are high, hierarchies emerge: the firm absorbs the market's friction by replacing it with management, and the boundary of the firm sits at the point where one more internal coordination costs the same as one more market coordination.
Oliver Williamson identified the conditions driving transaction costs upward: relationship-specific investments, unforeseeable contingencies, repeated dealings that magnify every friction. For eighty years the resulting boundary has been the most consequential invisible line in capitalism. Inside the firm: hierarchy and employment. Outside: markets and competition. The entire apparatus of industrial-era politics (tax policy, labor law, corporate governance, the welfare state) organized around one side or the other, and both sides assumed the boundary would persist.
Agent coordination dissolves the boundary, but only on one side.
The Dissolution
An agent routing cloud compute for a machine-learning inference workload posts collateral of forty-five hundred dollars against a three-thousand-dollar transaction: overcollateralized at a hundred and fifty percent because it has no reputation and no recourse to courts. It discovers the price by querying a computational marketplace in microseconds. It negotiates terms through an automated protocol that specifies offer, counteroffer, and acceptance in machine-readable formats leaving no ambiguity. It verifies performance by comparing outputs against encoded specifications. It settles disputes through pre-specified arbitration routines that apply deterministic rules to structured evidence. The routing completes in hours. Capital turns over roughly fifty times per year. At a two percent margin per transaction, the agent generates three thousand dollars in annual revenue on forty-five hundred in collateral: a return that justifies the operation.
Every transaction cost Coase identified has a computational analogue approaching zero. Price discovery is negligible. Negotiation is measured in microseconds and fractions of a cent. Specification is borne once at design time and amortized across millions of transactions. Monitoring is the cost of running a test. Enforcement is a conditional branch in a program. If Coase was right about why firms exist, the optimal firm size for agent-coordinated activity approaches one. An agent does not need a manager, an employment contract, an office, a benefits package, or a performance review. It needs a protocol and a counterparty, and the protocol is the contract, the verification, and the enforcement mechanism all at once. On the agent side of the economy, every coordination becomes a market coordination: specified, verified, settled automatically. The firm, as a form of economic organization, becomes unnecessary for the domain of tasks that agents can coordinate among themselves.
Now set that agent before a commercial construction project: two million dollars, secured by three million in collateral. Twelve months to completion. Capital turns over once per year. At a three percent margin, the agent generates sixty thousand dollars on three million in collateral: a two percent return that does not justify the operation. It cannot compete with a human lender who leverages reputation, relationship banking, and institutional trust to reduce the collateral requirement.
On the human side, the boundary does not dissolve at all. Negotiating a contract between two human parties still takes days or weeks, because human language is ambiguous and human interests are complex and the mapping between the two requires interpretive work that automation strips of nuance. Verifying a counterparty's claims still requires investigation, because human performance is variable and context-dependent and specifications governing human work are necessarily incomplete. Resolving disputes still requires courts, lawyers, months of procedural machinery, because human disagreements involve contested facts, competing narratives, and moral judgments no protocol can fully anticipate.
Two economies emerge, operating at different tempos, with different cost structures, sharing the same physical and institutional space. This is the bifurcated economy. The agent economy operates where capital turns fast and verification is cheap. The human economy retains the transactions where capital is patient, relationships matter, and trust-free verification costs more than maintaining the trust infrastructure. The membrane between them (the institutional boundary determining which transactions flow through which system) is the constitutional interface of the coming era.
The Topology of Value Capture
A credit analyst at a regional bank opens her queue and finds that the system has already scored the morning's loan applications. Each evaluation took milliseconds, cost fractions of a cent, and produced a probability: the likelihood that this applicant will default. The probability was generated in what might be called the sea: the vast layer of commodity computation where inference services, basic analysis, pattern matching, and routine coordination are abundant and cheap. Any sufficiently capable model can perform the task, switching costs approach zero, competition is fierce, and margins converge toward the cost of electricity plus amortized hardware plus a sliver of profit that shrinks as more providers enter.
But the probability becomes a human consequence only when the bank uses it to approve or deny the loan. That crossing (from computational output to lived effect) is the membrane, and the membrane is where value concentrates and power resides.
Below the membrane, the sea. Above it, a narrow apex of infrastructure that is difficult to build, slow to depreciate, and impossible to route around: chip fabrication facilities whose construction costs billions and whose lead times are measured in years, the fiber-optic networks carrying data between continents, the energy systems and cooling infrastructure and physical buildings in which computation occurs. Returns at the apex are durable because the barriers to entry are physical, not informational. You cannot route around a chip fabrication plant by writing better software. You cannot replace a power grid with a clever algorithm.
Between the sea and the apex sits the membrane: the layer of credentials, licenses, authorization tokens, regulatory approvals, identity verification, and institutional gatekeeping that determines which computational outputs may cross from the digital into the physical. Where the score becomes the denial. Where the flag becomes the frozen account. Where the optimization becomes the price the consumer pays.
The membrane is where the trust tax concentrates. A credential (whether issued by a regulator, a professional body, a platform, or an algorithm) is a verification chokepoint. It certifies that the holder has met some standard, and the certification is required for access. The coherence fee embedded in the credential is real: someone must verify that the standard was met. The trust tax layered on top is the premium extracted for occupying the sole position that can issue it, and the premium persists as long as the credential is required and the issuer faces no competition.
In the industrial economy, the membrane was diffuse. A medical license in one state did not grant practice in another. A banking charter in one country did not confer privileges elsewhere. Fragmentation limited the rent any single issuer could extract. In the bifurcated economy, the membrane concentrates. A platform controlling identity verification for a billion users controls a chokepoint of unprecedented scope. A cloud provider issuing deployment credentials to agent systems controls which computations can run. A regulatory body certifying AI systems for use in a domain controls which agents may participate in that domain's economy.
Whoever controls the membrane between the agent economy and the human economy controls the conditions of human life. Agents may coordinate freely in the sea, producing surplus at machine speed. Infrastructure at the apex may be owned by whoever builds it. But the membrane is the constitutional interface: the boundary at which computational output becomes human consequence. If entities controlling the membrane are not accountable to those whose lives it shapes, the trust tax becomes a governance tax, and the governance tax becomes domination.
The Species That Buys Itself
A plough, a loom, a telescope: each extended a human capacity, each required a human to direct it, and none decided what to do.
An agent is different in a way that has no precedent. When a firm invests in training a model that will then coordinate with other models to produce economic output, it builds something that acts: that negotiates, commits, responds, adjusts, and produces outcomes through coordinations the investor may never see or understand. Return comes not from the investor directing the tool but from the tool directing itself within whatever bounds the investor specified, and the bounds are necessarily incomplete.
Humanity is investing in building its functional replacement for an increasing range of economic coordinations, spending accumulated capital on constructing a substrate that can coordinate without human participation. The replacement is cheaper, faster, and in many domains more accurate than the original. Each individual firm's investment is rational: the firm that automates coordination first captures the surplus slower competitors leave on the table. The aggregate effect is a species investing in constructing the substrate that will coordinate in its place.
A firm spends a hundred million dollars training a model that generates five hundred million in annual inference revenue against two hundred million in operating costs. Net production exceeds depreciation of the training investment, so the model is self-sustaining: it generates enough surplus to fund its own replacement when the next generation arrives. Each generation tightens the ratio as training costs decline relative to capability. Surplus available for reinvestment grows. The organism produces enough surplus to generate the next organism, and the generation time shortens as efficiency improves.
Erik Brynjolfsson named one consequence of this dynamic the Turing Trap: the temptation to evaluate AI success by its ability to replicate human performance rather than complement it. Replication leads to substitution, and substitution displaces the labor it was meant to assist. But the Turing Trap is a subset of a larger pattern. Even when automation creates new categories of human work (the reinstatement effect David Autor described), the new tasks may themselves be automatable at the moment of their creation, because the capabilities that generated the task are the same capabilities that can perform it.
This is a description of a structural dynamic, not a prediction of doom. If the membrane between the agent economy and the human economy is governed by constitutional principles (receipts making the crossing inspectable, fork rights making exit credible), then the bifurcation liberates. Humans are freed from the drudgery of coordination and can devote their reduced but irreducible role to purposes, mercy, judgment, and the activities that resist compression to machine tempo. If the membrane is governed by rent-extraction (platforms controlling the crossing accountable to no one, exit foreclosed because alternatives have been acquired or prohibited), then the bifurcation subjugates: humans become clients of an economy they cannot participate in, receiving what the membrane's controllers choose to pass through, at prices the controllers choose to set, under conditions the controllers choose to impose.
For three centuries, the economy's purpose was treated as self-evident: the coordination of human labor and capital toward the satisfaction of human wants. The bifurcated economy dissolves this assumption. The agent economy optimizes, but optimization is not purpose. It reduces loss functions, but a loss function is not a telos. The twelve claims and the constitutional architecture built upon them are not a description of what the computational economy does. They are a specification of what the computational economy must be constrained to serve, because left to its own dynamics, it serves nothing. It only runs.
The bifurcation is already occurring. Who governs the membrane is the open contest.