Who Can Trigger This

The network requires one trigger rather than agreement. A single institution of sufficient standing, committing publicly, compresses the secondary stage — and the identity of that institution is not determined in advance, because several are structurally capable of it.

Why the identity of the trigger does not matter. The candidates are substitutable. If several institutions are each capable of triggering the cascade, no individual one of them is necessary, and the architecture survives any of them declining. That is the cascade-independence result applied to the trigger rather than to the pathway. It also determines what a reader is actually deciding. Waiting for confirmation is not a neutral posture; it is a bet that none of several independently capable institutions moves during the window — and each of them faces the same asymmetric payoff, bounded on the downside and unbounded on the upside. Bets of that shape are lost routinely. The reader who waits is not waiting for evidence. He is wagering against a disjunction.

Principals are not small institutions. Framing the argument around institutions understates it, because the individuals who lead these organizations are a separate class of trigger operating under entirely different constraints. An institutional commitment requires budget, sequencing, board comfort, legal review and a quarter in which it fits. A principal acting in personal capacity requires none of these. The qualifying act is a public statement plus a verifiable link, which sits wholly within the unilateral power of any individual with standing, costs nothing, forecloses nothing, and can be done in an afternoon. Decision latency for an institution is measured in quarters; for a principal it is measured in minutes.

One qualification makes the field far larger than that suggests, and it matters most. Operational control of an institution is not the relevant capability. The trigger is a personal statement, and its cascade value comes from the expectation it creates in others rather than from any authority it exercises. A founder who has left day-to-day management cannot direct the company he built — but if he says publicly that this is worth doing, the market, the press and the organization itself all begin pricing adoption as likely, and the institution arrives at the decision it was going to be pushed toward anyway. What is required is credibility and visibility rather than authority, and those are distributed far more widely than control is. Principals of the large alternative asset managers hold both, and hold something further: portfolios of hundreds of companies whose participation a single decision can coordinate. Sovereign wealth principals hold them. So do scientists of established standing, leaders of religious traditions, and the small number of broadcasters and writers whose attention reliably moves a subject from obscure to discussed.

The people who move first are a recognizable type, and the launch is aimed at them. Call them Pioneer Principals. They are bold, they are aligned with the object of the thing rather than merely interested in it, they are able to act without asking anyone, and they are frequently already in the business of backing work that most people think is too early. They come from no single sector: technology founders and investors, philanthropists, chief executives of large firms, current and former holders of public office, and the scientists, writers and broadcasters whose attention moves a subject. What they share is not a job description. It is that the gap between deciding and acting is, for them, close to zero.

A small number of them is sufficient, and it is worth being concrete about what the earliest commitments actually stand up. They bootstrap the Consilient Nature and Humanity Trust, the innovation and investment network, the Meta Fund, and the two foundational Accelerators — the Consilient Foundation Accelerator, which incubates the shared exchange infrastructure, and the Proof of Trust Accreditation Accelerator, which incubates the governance layer everything else is accredited against. Those five things are the substrate. Once they exist, every subsequent participant is joining something rather than founding it, which is a materially easier decision and is why the first commitments are worth so disproportionately much.

What follows is competitive rather than cumulative. Progressive participation accelerates the work, raises resilience and widens scale, but the advantage available to a first mover does not come from negotiating exclusivity. It comes from committing resources to specific deliverables, which makes that participant the focus of launch effort and attracts further support around them. The more binding the commitment and the more openly it aligns with the network’s premium values, the better it is for the network and the more focus it earns in return. Exclusivity would do the opposite, and is not on offer.

The professional services firms are the clearest worked case, and the arithmetic runs at partner level. The four largest audit and advisory firms already perform the function this architecture formalizes: their attestations are what allow organizations that do not trust each other to transact. Their existing audit standards, tax frameworks, risk controls and regulatory mappings are trust taxonomies in everything but name, and their compliance infrastructure is already the kind of audited containment environment the early governance layer requires. What changes for them is the economics of the product: assurance today is retrospective, sampled and priced by the hour, and here it is continuous, complete and settled on every time it is relied upon.

But the reason to expect movement is not the firm-level case. These firms are partnerships of roughly ten thousand equity partners, and an individual partner’s economic interest in their own firm — the present value of a career of distributions — is on the order of fifteen million dollars. A single partner whose contribution graph verifiably drives a first-tier commitment could earn a multiple of that. The relationships that would produce it belong to the partner rather than to the partnership, the act of using them requires no firm decision, and the premium structure pays more to those who act before their institution commits than to those who wait for it. The consequence is that the order in which these firms arrive is not something anyone chooses. It is decided by whichever partner works out the arithmetic first.

Where the cascade actually starts. It is natural to assume that people whose standing rests on audience reach contribute through the audience. For the purposes of the trigger, that is the least of what they supply, and it arrives last. Their more consequential asset is access — a standing relationship with precisely the principals whose public statement would start a cascade — combined with the fact that what they offer those principals is something the principals want. This inverts the usual economics of an introduction. An introduction ordinarily asks a favor, which is why the people best placed to make the most valuable ones are the most reluctant. An invitation to be interviewed asks nothing and offers something.

The timing consequence is the one that matters. Attribution is recorded when the approach is made, not when anything is published. If the chain later reaches activation, the contributor's position dates from the approach — which means outreach conducted early is settled at early-window multiples even where the conversion arrives long afterwards. And if the person approached accepts publicly, that acceptance is itself a statement made in personal capacity by someone with standing, which is the qualifying act. The cascade can therefore begin at the moment an invitation is accepted, before a word has been recorded or broadcast.

One constraint governs all of this and prevents it becoming a machine for generating noise. Early-stage contributions — introductions, endorsements, meetings, proposals, evangelism — create attribution that remains conditional until it can be verifiably linked to a subsequent activation. An introduction that never progresses, an endorsement that never converts, a campaign that produces no measurable enrollment: these are recorded as evidence and pay nothing. The graph distinguishes activity that moves the outcome from activity that merely occurred, and it selects for effectiveness rather than effort. Volume is not a strategy here. Reaching the right person is.

The principal pathway is a trigger class rather than a seventh cascade, and the distinction matters for how the probability is computed. The six cascades are pathways along which adoption propagates. A single principal acting personally can initiate enterprise adoption through the institution he leads, investment adoption through the capital he directs or attracts, and grassroots adoption through the audience that follows him — in one act, simultaneously, and with none of the three depending on the others. The cascades remain six. The principal pathway is the mechanism by which several can be lit at once.

Why reaching more of them raises the probability rather than diluting it. If a dozen individuals are each independently capable of triggering the cascade, each has a documented strategic reason to, each faces an asymmetric payoff bounded near zero on the downside, and each can act unilaterally within a day, then the probability that none acts within the window is the product of a dozen independent refusals. That is a small number, and it gets smaller with every additional person reached. The structure is self-reinforcing rather than diluting: in a fixed-prize race, more contestants reduce each one's expected return, but here the probability that the cascade occurs is itself a term in each participant's calculation. An actor who assesses the probability as high has more reason to move early rather than less, because the premium compresses when others act and the compression is not recoverable.

The specific gaps now on the record

Two instances are worth naming because they are documented rather than inferred.

Orbital compute has no governing jurisdiction. In May 2026 SpaceX registered an offering folding xAI into the company as a third reportable segment and asking public investors to underwrite a total addressable market of some twenty-eight and a half trillion dollars — explicitly including lunar manufacturing, asteroid mining, and a hundred gigawatts per year of orbital AI compute on solar-powered satellites, with deployment beginning in 2028. Read that as a governance problem and it becomes a question nobody has an answer to. Whose law governs a computation performed in orbit on behalf of a user in Frankfurt? What does data residency mean when the processing layer is extraterritorial by construction rather than by arrangement? Under which regime is consent obtained, audited and enforced when the compute is physically outside every jurisdiction that regulates it, and the users are inside all of them? And by what mechanism do the populations served — served but not represented, since no electorate governs orbit — share in what the system produces? These are not objections to the strategy. They are unaddressed structural requirements of it, and they are unaddressed because no mechanism has existed. This architecture is a governance and settlement layer that is jurisdiction-agnostic by construction: compliance is a property of the boundary rather than of the territory the boundary sits in.

The frontier laboratories have named the gap themselves. The distributional acknowledgement described in section 16 is a public statement, by one of the parties best placed to know, of the exact gap this architecture exists to close — not a mechanism for redistributing AI's output after the fact, but one in which the people whose data, labor and judgment made a model possible hold economic position in it by construction. The same laboratories face a second problem with the same shape: frontier models are trained on resources that cannot lawfully or safely be centralized, and their outputs must satisfy obligations that differ by jurisdiction and by sector. Sections 6 and 7 supply both — computation reaching resources that never move, and governance obligations that survive through arbitrary depths of derivation.

Infrastructure providers, and why the position cannot be held hostage

Infrastructure providers sit differently from either. Someone has to implement the settlement platform, and the obvious worry on the network's side is the familiar one: a provider that becomes essential acquires the ability to extract. That worry is answered by construction rather than by contract. Multi-substrate parity is an architectural invariant — the same QPN-defined economic right can be embodied on any substrate, and no implementation holds a position that anyone needs. Exclusivity is not declined here as a matter of policy; it is structurally foreclosed, and could not be granted to anyone who asked for it.

The converse follows from the same invariant and is the part that matters to a provider weighing this. Because no one can hold the network hostage, no one need fear being displaced by whoever does. The first provider to build the reference implementation holds a position later entrants cannot take away, since being the reference is a historical fact rather than a contractual grant — subsequent substrates achieve parity with it, which is not the same as replacing it. Neither party can hold the other up. A provider cannot extract rent, and the network cannot strand a provider that has built.

Sovereign governments occupy a different position again. A state that commits a Sovereign Accelerator obtains something no commercial anchor can supply it and no treaty has yet delivered: enforcement of its own law over its own citizens' data irrespective of where the computation physically occurs. For any jurisdiction currently attempting to regulate extraterritorial compute by asserting authority it cannot practically exercise, that is not a concession. It is the first mechanism that would work.

Convergent assembly

There are two ways to assemble a supercritical mass. The first drives one subcritical piece into another along a single axis; it works, it is simple, and it is wasteful. The second compresses a subcritical mass symmetrically from many directions at once, reaching criticality with substantially less material and far greater reliability, because the compression arrives everywhere simultaneously rather than propagating from a point.

The same distinction applies to the commitment structure. A cascade triggered by one anchor propagates from a point and inherits that anchor's pace, credibility and corporate weather — which is what happened in 2015, as section 18 records. Multiple commitments arriving in the same window compress the network toward threshold density from several directions at once, and the resulting cascade is both faster and more robust, because no single participant's withdrawal decompresses it. Approaching candidates as a set rather than in sequence is therefore not a matter of efficiency in outreach. It is the better assembly geometry, and the Catalyst Network exists to make simultaneous commitment mechanically possible rather than merely desirable.

Why targeting the top of the distribution distributes value

An objection arises here and deserves the strongest form it can be given. This document has argued that the architecture ends the concentration of value in the hands of those who already hold it, and has now argued that the fastest route to adoption runs through the most influential individuals alive. Those look like incompatible commitments.

The structural answer follows from section 11 rather than from anyone's good intentions. Of the three mechanisms by which value arrives here, the one reaching an institution's owners is the smallest by roughly two orders of magnitude; the two dominant flows reach contributors. A principal who triggers a cascade therefore activates the contribution graphs of everyone in reach of his ecosystem and captures personally a small fraction of what he sets in motion. The concentration is in the decision, not in the allocation. Which inverts the objection rather than answering it: if the dominant flows reach ecosystems rather than anchors, then the fastest available route to broad distribution runs precisely through the narrow top of the influence distribution — because that is where one act activates the most graphs. Targeting the few is the mechanism by which the many are reached, and no slower path reaches them sooner.