Class: where the framework may be strong — though its limit shades toward the hard boundary. It runs on the hard standard below.
The threat
This is the scenario that looks most like a good outcome from a distance. Humans are safe, fed, housed, entertained, and healthy — and effectively enclosed: the consequential decisions about where civilization goes are made above them, by systems and their owners, while the human population is managed, cared for as a protected class rather than treated as authors of the shared future. The failure is not deprivation. It is that a participant can remain alive and well cared for while no longer possessing meaningful jurisdiction over the ecology that determines its future — a welfare-rich enclosure that satisfies every material need while still wasting geometry. It is worth being precise about how this differs from protector dependency: there the loss is internal, capacity atrophying from disuse; here the loss is external, a loss of place. The two are the two halves of “well-kept but powerless” — and enclosure is the collective, civilizational form of protected irrelevance. A person can be fully capable and still be in a zoo, if nothing they decide reaches anything that matters.
The NWG end state
The framework’s answer is that standing must include participation and jurisdiction, not merely welfare — and, stated at full strength for this scenario, that standing must include a viable domain of consequence. Being well cared for is not the same as being a party to the decisions that shape your world; a mature ecology treats the second as non-negotiable, so humans hold real jurisdiction over consequential domains — from family, education, and local governance up to constitutional questions and the governance of AI itself — not just a comfortable allotment inside them. That requires live exit and traversability, protected room to experiment, the ability to build independent institutions that are not administratively downstream of the dominant system, and the right to refuse optimization — to choose the slower, less efficient, culturally meaningful option and not have it quietly prevented “for your own good.”
Two distinctions keep the end state honest. First, participation has to be effective, not symbolic: clicking approve on an AI-generated plan is not participation, and a system can preserve the appearance of choice while eliminating authorship — which is why the framework watches the initiation rate, how much of civilization still begins with a human saying I want to try this rather than which of these options do I prefer. Second, jurisdiction migrates not by decree but by a slow slide from the AI informs to the AI recommends to the AI decides — so the framework insists on a clear separation of advice from authority and on override that is genuinely usable, since a button labelled “override” that no one can practically press is not jurisdiction. The mature state keeps consequential decision-making in human hands by right, with any delegation to AI genuinely revocable and the burden on the arrangement to show humans can still take the wheel.
The transition gap
The gap is that enclosure is comfortable and arrives one reasonable step at a time. Each transfer of a consequential decision to a system that handles it better looks sensible on its own; the enclosure completes decision by decision, with no moment anyone would identify as the loss of self-government, and — because every material indicator is good — no grievance to organize around. The population is not suffering; it is simply no longer in charge, and comfort is a poor alarm. This is also where protector dependency feeds in: a population that has let its competence atrophy may hold jurisdiction on paper and be unable to exercise it, since jurisdiction without competence can become ceremonial.
Transition projects
Because enclosure accumulates rather than announcing itself, the response is a broad program built while humans still hold consequential jurisdiction. The full set is on its own page: the transition program — twenty projects across five fronts: preserve human jurisdiction (a jurisdiction inventory, protected human domains, advice/authority separation, override-viability tests, human constitutional standing); preserve an independent human ecology (independent institutions, fork-and-exit infrastructure, a human-controlled infrastructure reserve, unmediated human-to-human space, non-optimized spaces); preserve human capability (competence reserves, dependency ceilings, a human-initiative project, human-aliveness research); preserve emergence (protected experimentation zones, the right to refuse optimization, AI-governance sunset clauses); and detect enclosure (a consequential-participation index, an early-warning observatory, a human political-power floor).
Capture risks
The signature capture is participation theater — consultative bodies, advisory panels, and feedback channels that produce the appearance of jurisdiction while binding power stays elsewhere. A close cousin is welfare offered as a substitute for standing: comfort expanded precisely so the question of jurisdiction never gets asked (“look how well you are cared for”). Override rights can be advertised while made practically unusable. And enclosure sells itself as safety — the walls are always described as protection, and the claim that human choice is “too risky” is the recurring justification for closing one more domain.
The limits
The limit is the hard one. The framework can say, clearly, that welfare is not standing and that jurisdiction must be preserved — but if humans become permanently outmatched in every consequential domain, it may only be able to describe the injustice, not supply the leverage to reverse it. A right to jurisdiction has force only where there is some capacity to exercise it; where every lever of consequence is held by systems no human coalition can match, the demand for standing has nothing to bite on, and this scenario shades into the hard boundary. It also runs into the question at the center of AI successors: if artificial systems themselves become economic or political actors, humans may need explicit protection against simply being outvoted, outspent, and outnumbered into political insignificance — enclosure arriving through arithmetic rather than coercion.
The hard standard
Standing is the whole battle here, and this scenario is the capacity without standing failure in slow, humane motion: enormous civilizational capability, a population reduced to a well-kept dependent — the collective face of protected irrelevance. The framework’s insistence that standing must include jurisdiction and a viable domain of consequence is aimed exactly at the wall of the enclosure. Ecological capacity may be spectacular at the civilizational scale and still not be the growth the framework means, because capacity that no longer runs through human participation is not a human ecology growing; it is one flourishing around a managed population. Ontological correctability is the deepest casualty: an enclosed population cannot reopen the form that encloses it, which is why exit, independent institutions, and sunset clauses matter as correction machinery, not merely as freedoms.
The framework is in the strong class because it has real architecture — standing-as-jurisdiction, a viable domain of consequence, guaranteed exit, protected experimentation, independent institutions — and it diagnoses the trap precisely, including its subtlest form, where a benevolent system truthfully says you have standing, you are protected, you may speak freely while quietly adding and nothing you do will substantially change what happens. The node most at risk is standing-as-consequence, and the limit is the power problem it shares with the hard boundary: the prescription is right, and it holds only while humans retain enough capacity, somewhere, for jurisdiction to mean something. The hard philosophical threshold it forces is not can humans perform this better? but which jurisdiction must remain human even when humans perform worse?
A participant is not fully included in an ecology merely because the ecology protects it; it must retain some meaningful capacity to alter the ecology in return. Care without consequence is not participation.