Every argument in the last three chapters rested on one assumption, and it is time to stop assuming it and defend it. The assumption is that inside this people, power accumulates nowhere - that the vote cannot be bought or multiplied, that no office hoards authority, that neutrality is protected by structure and not by the good character of whoever happens to be in charge. If that is false, the whole edifice is a sermon. And the objection that presses hardest, the one raised by every serious person who has watched a movement curdle into a hierarchy, is exactly this: you cannot gather a mass into an actor without, sooner or later, breeding a caste inside it. Someone ends up steering. The many become the led. This chapter is the answer, and the answer is not a promise of virtue. It is a set of mechanisms, each of which can be inspected.
The objection deserves to be taken at full strength, because it is not idle pessimism; it is close to a historical law. Study the life cycle of movements and parties and unions, and the same arc recurs with grim regularity: a body founded to give the many a voice develops, out of the sheer practical need for someone to run things day to day, a stratum of organizers who know the machinery - and that stratum, over time, hardens into an interest of its own, distinct from and eventually opposed to the members it was meant to serve. Those who administer the organization come to own it; the administered become, once again, the led. This is the reef on which nearly every attempt at collective self-rule has broken, and a project that waves it away has not understood the problem it claims to solve. What follows is offered against exactly this, and it is worth nothing if it cannot meet it.
Start with the vote, because everything rests on it. In this people, one verified human is one vote, and the vote is bound to the personhood credential the earlier chapters described. This does two things at once that no ordinary system does together. Because each vote is tied to a verified unique person, no one can manufacture a majority out of invented accounts - the Sybil attack that hollows out every anonymous space is simply closed. And because standing attaches to the person and to nothing else, it cannot be accumulated: reputation does not convert into extra votes, contribution does not, and the internal unit of economic exchange does not. The wealthiest member and the newest count exactly once, each. Whatever else concentrates in this world - and much does - the one thing that cannot is the vote.
This is a sharper choice than it looks, and it is worth marking the road not taken. There is a sophisticated tradition, much of it native to the same technical world whose tools this book praised, that tries to improve on raw majority rule by weighting votes cleverly - letting people express how strongly they feel, at a rising cost, so that intensity registers and the bare headcount does not tyrannize. It is elegant work. But every such scheme, however clever, still weights: it lets some signal other than the equal human count modulate the outcome, and wherever there is weighting, resources find their way in, because the ability to spend more - of tokens, of credits, of whatever the currency of intensity is - never distributes evenly. This people refuses the weighting entirely. It does not try to be clever about how much a vote is worth. It makes every vote worth the same, and puts its intelligence somewhere else.
Because raw equality has a real defect, and it has to be met rather than ignored: a bare majority can be uninformed, hasty, or cruel to a minority, and "one person, one vote" alone does nothing to prevent it. The answer here is not to weight the vote but to structure the road it travels. A decision of consequence does not go straight to a show of hands. It passes through stages: a proposal must be stated with its rationale and its risks; it undergoes review by those with relevant competence; it is opened to deliberation, where anyone may question and amend it; it sits through a cooling-off period, so that a passing heat does not decide a lasting thing; and on specialized or high-stakes questions the threshold for passage rises, and the reasoned objection of a competent council can lift the bar that consent must clear. Note what this does and does not do. It does not let the experts decide - they cannot override the vote, only inform it and slow it. And it does not let a momentary majority ram a thing through. Quality is loaded onto the process; equality is preserved at the ballot. It is protection against the mob without surrender to a priesthood - populism refused, and epistocracy refused with it.
It is worth seeing how sharply this breaks the familiar geometry. Every model of common life we know - historical or modern - is built on the same pyramidal scheme: at the top a king, a chief, a president, a general secretary. The apex is at once the entry point and the exit point of every large decision - a narrow neck through which only a few can pass, and almost always the same few: the entourage, the court, the circle. Titles and rituals change; the geometry does not. What is proposed here is a different geometry. The leader is not limited or checked - the leader is absent: not a vacancy waiting to be filled, but a variable the equation does not contain. This is laid down in the founding documents as a principle and held in place by the architecture as a fact: each member self-governing, the whole self-organizing, with no apex at all.
And there is, by design, no seat for a caste to occupy. There are no parties, no standing officials, no executive that governs between votes and slowly becomes the government. The functions that must exist - keeping the software running, acting in a genuine emergency - are defined as service, not command: narrow in scope, capped in power, rotated by design, and unable to touch the substance of what the people decides. Power is not concentrated and then checked; it is prevented from concentrating in the first place, which is the stronger thing, because a check can be captured and an absence cannot. Rotation; the separation of roles, so that no one holds several levers at once; the mandatory declaration of conflicts of interest; the protection of minorities; the right of any member to put a question on the table from below - these are not adornments. They are the plumbing that keeps the caste from forming.
But I promised mechanisms that can be inspected, not a design that asks for faith - and no design is perfect, so an honest architecture needs a last resort for the day the rest fails. It has one, and it is the most important safeguard of all, precisely because it does not depend on the good behavior of anyone inside. The whole thing runs on open code, and the members hold the right to exit and to reproduce it. If it were ever captured - by a founder who would not let go, by a faction, by a buyer - the members could take the entire structure, the code and the rules and their own verified selves, and continue it elsewhere without the captor.
And here is why that is not mere disruption: a hostile fork can copy the code, but it cannot copy the people, the legitimacy, or the verified membership, which travel with the members and not with whoever holds the keys. A legitimate continuation is the one that keeps the founding core intact and carries the majority of the verified people with it; a seizure keeps the buildings and loses the city. This is what makes "trust me" unnecessary. The founder's power is arranged so that it can only ever narrow, and if he betrays even that, the people can walk out the door with everything that mattered and leave him holding an empty copy. A structure you can leave, taking its whole worth with you, is a structure that cannot, in the end, be captured.
I will not oversell even this. It is a design, and designs meet reality and bruise. Money is endlessly inventive about finding new doors; some of these mechanisms have been tested only at small scale; and the right to fork is a backstop precisely because the front-line defenses are not guaranteed to hold. That the architecture makes capture hard is a claim I can defend. That it makes capture impossible is not, and I would distrust anyone who told you otherwise. But "hard, with a real exit if it fails" is a great deal more than any hierarchy offers, and it is the honest most that can be promised.
Which brings the book, at last, to the place it has been pointing all along: an honest accounting of where the whole thing is weak. Not the governance alone, but the entire argument - what it still cannot do, what could break it, and the plain fact of its present size. A book that has asked, all the way through, to be tested at its weakest joints rather than admired at its strongest owes a final chapter that does exactly that.