The Call
What follows is for the living to decide.
Agent systems coordinate through objectives they did not choose. Nothing in the act of optimization supplies a common good.
An objective function is a constraint imposed from outside. The system that denies a loan while its applicant sleeps contains no account of why she needed the money, what the denial will cost her, or which reasons should matter more. It returns the result its designers made operational. Cruelty or kindness enters through the human purposes, institutions, and consequences around that result, including the choice to treat it as final.
Humans learn ends differently. We learn them under finitude: time that will run out, lineage that binds us to the dead and the unborn, responsibility that accrues because mistakes have faces. A face can be wronged. A wrong can be remembered. We inherit purposes before we choose them, and even when we try to escape the weight of choosing, it follows us. A child inherits a language, a history, a set of debts and obligations that precede her birth and will outlast her death. A community inherits its institutions, its customs, its unresolved conflicts, its accumulated wisdom and accumulated error. Inheritance is the condition under which morality exists. A being that lacks the capacity to suffer consequences lacks the capacity to bear obligations. A system that processes without stakes cannot betray, because it cannot promise.
The book's conjecture is that we are spandrel souls: consciousness arrived with the biological machinery once required for coordination, and coordination can now migrate to systems for which no claim of felt experience is needed. The dome no longer needs the mosaics. The mosaics still weep. Whether or not the conjecture survives, the constitutional argument does not depend on it. Efficient allocation matters because it improves or diminishes lived human lives, and those consequences create claims against the institutions that deploy the system.
The asymmetry between people who bear consequences and systems that score them is a political fact. The answer proposed here requires durable witnesses, disclosed costs, contestable receipts, portable standing, and an institution willing to limit a true record's reach. Some of those institutions do not yet exist. They will need money, law, training, and people prepared to answer for their judgments.
If systems without legal or material stakes wield power over lives that end, safeguards cannot rest on the virtue of the process. They belong in the surrounding institutions: receipts that make coercion legible, capabilities that can be revoked, exits that remain practicable, and records that survive hostile interpretation. The system will not supply those constraints for us. People with standing must build and govern them.
The argument has reached its practical question: whether the architecture will be built, and against what resistance.
Two futures remain possible. They differ in what power can be made to answer.
In the first, constraint arrives without a face. Scores shift while you sleep. Associations become liabilities by correlation. Distance is "recommended," and the recommendation is priced into every rail that matters. No one threatens. No one argues. Refusal is made expensive in ways untraceable to a contestable act. The deepest luxury the lords sell is not convenience. It is abdication: the right to stop being the author of one's life while still feeling free.
The exile adapts. She learns to navigate around invisible boundaries, to cultivate the favor of systems she has no window into. She becomes, without anyone calling it this, a subject.
In the second, constraint is never silent. When a system limits a person, it leaves a receipt. The receipt names the act, the authority, the bounds, the justification, the path of appeal. Not every receipt is read; not every act is contested. What changes is standing: the governed can verify what governs them. The exile holds a receipt. She may still lose. But she has standing: proof that something happened, to her, by a nameable process. She can contest.
Coercive authority becomes inspectable by those subject to it, while private life is no longer made legible by default to systems that will never answer for what they know. Politics does not disappear under that arrangement. What disappears is the silent exercise of power.
A skeptic who has followed this argument might raise a final objection: "You have shown what is possible. You have not shown what is likely. The Quiet Foreclosure is stable because it serves the interests of those who control it. The Protocol Republic requires overturning an equilibrium that no one who benefits from it has reason to overturn. Elena's second story is a fable. Her first story is the future."
The objection deserves a real answer rather than an inspirational one.
The Quiet Foreclosure is stable. Those who benefit from it are powerful. Those who suffer from it are dispersed, unaware, and poorly organized. This book's proposed transition is not inevitable, nor even likely if likelihood is measured by the current distribution of power and interest. Platforms that control identity, payment, and reputation have market capitalizations in the trillions, lobbying budgets in the billions, and user bases trained to prefer convenience over sovereignty. Every entity with a material interest in the status quo will resist.
One structural difference distinguishes this moment from earlier cycles of decentralization and recapture. The internet's original architecture was captured because it provided connectivity without economic commitment. Anyone could route packets, but no one staked value or bonded performance at the protocol layer. What coordination required above bare connectivity (identity, payment, trust, recourse) was supplied by platforms, and supply conferred sovereignty. The verification primitive changes the equation. For the first time, economic commitment can be embedded in the protocol itself: staked collateral, bonded arbitration, receipted authority. Whether this advantage proves durable is an empirical question. That it exists at all is why the possibility deserves to be taken seriously.
Equilibria also shift for reasons the entities benefiting from the equilibrium are powerless to control. Costs of the current architecture are becoming visible: not through philosophical argument but through accumulation, each breach that exposes millions of records, each wrongful freeze that destroys a livelihood, each Elena. The accumulation is slow, but it is also structural: the same computational systems that make the Quiet Foreclosure possible make its failures legible in ways that earlier forms of institutional abuse were not. A wrongful freeze in 1990 was a private grievance between a customer and a bank. A wrongful freeze in 2025 is a social media post, a news story, a class action, a regulatory inquiry. Visibility does not guarantee change, but it creates conditions under which change becomes politically possible.
History offers no guarantees. The Roman Republic lasted five centuries before the principate captured it. Venice lasted a millennium before Napoleon dissolved it in an afternoon. The American Republic has lasted two and a half centuries and is not obviously secure. Constitutional constraints that limited power in one era were captured or abandoned in the next. Design does not guarantee outcome.
But design shapes the possibility space. A world where the Protocol Republic is possible is different from a world where it is not. A world where Elena can hold a receipt is different from a world where she cannot. Architecture does not determine politics, but it constrains what politics can achieve.
The architecture can be written down. It does not build itself, and the interval in which it can be built is passing rather than theoretical.
A Protocol Republic must be built in code that refuses custodial shortcuts, mechanisms that bind power to receipts, credentials that remain portable under pressure, and communities willing to bear some friction to avoid domination. The institutional work is larger than the technical work. Arbiters need training, funding, and a professional culture of impartiality. Independent rulemaking bodies need charters and oversight. Interoperable protocols need standards, reference implementations, and incentives that make compliance more attractive than proprietary lock-in.
The easier order will win unless opposed. Its bargain is simple: accept the terms, surrender the keys, trade agency for relief. It asks almost nothing and delivers comfort immediately. The alternative asks for competence, attention, and the willingness to be inconvenienced by one's own freedom. Freedom is more expensive than servitude. It requires the citizen to attend to matters the subject can safely ignore: who makes the rules, how the rules are enforced, whether enforcement is proportionate, whether appeal is genuine. Systems that offer epistemic foreclosure, the surrender of judgment for relief, are not lying about the convenience. They are lying about the cost, which is not assessed in the present but compounded in the future, when the institutions that could have constrained power have atrophied from disuse and the power that could have been constrained has grown beyond the reach of any mechanism the citizen might then wish to invoke.
The consent that authorized the arrangement is itself a casualty. Terms accepted in 2019 by a person who no longer exists, under conditions that no longer obtain, amended unilaterally as the original agreement permitted: the agreement is dead, and the platform continues to animate its corpse. This is necromantic consent: invoking a dead agreement to authorize present coercion. Article 10's anti-waiver principle exists because contractual consent in this domain has become a simulacrum: the form of agreement performing the function of agreement while lacking its essential feature, the ongoing and informed participation of a living person.
One diagnostic governs every encounter with computational authority:
What trace will remain?
If the answer is none, domination is present, whatever its manners.
If the answer is a receipt only power can read, accountability is decorative.
If the answer is a receipt verifiable by affected parties, a different order is possible. Not guaranteed. Possible.
Verification cost is the primitive variable that determines the structure of trust, and therefore the structure of power. No economic shift since the collapse of communication cost rivals the collapse of verification cost in constitutional consequence. The printing press made it cheap to disseminate information, and the constitutional response took three centuries to mature. Cheap verification makes it possible to prove claims without trusting the claimant, and the constitutional response has barely begun.
Power will be computational. Verification will be cheap.
The question is whether the receipts will belong to those who govern, or to those who are governed.
Demand receipts from computational authorities, preserve a penumbra where private life can remain unobserved, and give personal records an end. The proof of power must endure; the proof against persons must expire.
The clerk's pen still moves. The ledger still updates. In one future, the entry vanishes without trace, and the erased learn their status by collision with a world that has already decided against them. In the other, every line leaves a mark that outlives its author, and the mark can be read by those it concerns.
The question is which we will build.
Freedom needs receipts.
Humanity needs mercy.