A World of Protocols.

What is a protocol?

An engineer's take on voluntary human cooperation.

A protocol, in the engineering sense, is a set of rules and a working implementation of them. TCP/IP is one. HTTP is one. So is the system that clears bank settlements overnight, the system that hands email between two providers, and the system that lets two strangers' wallets agree on what a Bitcoin is. None of these is owned by any one company or state; each is a piece of neutral infrastructure that does work nobody has to be in charge of.

People also live inside dozens of social protocols — a marriage, a chess club, a hospital, a factory floor, an open-source project, a religious congregation — each with its own rules about participation, decisions, and how to leave. Those run on trust and habit, and they cap out at the size where trust and habit can still carry them.

This page is about applying the engineering shape of the first kind — neutral, implementable infrastructure for cooperation between strangers — to the work the second kind already does, at scales habit alone can't reach. The proposal is concrete: a working attribution architecture — machinery that decides who did what and what it is worth, and routes value accordingly. It runs the pricing, holds the reputation, mediates disputes, runs votes, and bridges value and identity between protocols. People do the work; the system makes the cooperation possible where informal trust would have broken down.

Call this, broadly, a world of protocols: many of them, each with its own values, mission, and culture, with people belonging to several at once the way a person can already be a citizen of a country, a member of a profession, a parishioner of a church, and an employee of a company all at the same time.

Rules, work, posture

Look at it from three angles: rules, work, posture.

The rules mostly live in pricing — how contributions get valued, how they get paid, how that payment changes over time. A small, deliberately stable set of foundational rules lives in a constitution, where the things pricing cannot decide get decided: what the protocol exists for, what it refuses to host, what minimum care it provides for those who cannot contribute, how it treats people outside its membership.

The work is unusually broadly defined. Producing something measurable is work, but so is mentoring someone, brokering an introduction, refusing to onboard a bad actor, or mediating a conflict before it ruptured. Standing ready as a firefighter or paramedic for a year in which nothing happens is work, and so is drafting and arguing about the rules themselves. The central claim is that all of these are real, all can be valued, and all should be paid.

The posture is voluntary, always. People join because the protocol serves them, leave when it doesn't, and routinely belong to several at once. The protocol claims no sovereign authority and enforces nothing by violence. It lives inside the institutions you already live inside, paying its taxes and signing its contracts and obeying its laws, competing for your time on the only basis a voluntary association ever has: actually being worth your while.

Honest in time and space

The architecture is honest in two dimensions where current institutions are blind: time and space.

It is honest across time. Contributions get a best-guess price now and are re-priced over years and decades as their effects materialize. The teacher's lessons are paid in her seventies, when the work of the cohort she taught has been re-priced upward and the credit traced back to her. The maintainer's quiet years are paid once it becomes clear what their absence would have cost. Long causal chains stop being the thing institutions cannot see.

It is honest across space. Contribution credit, reputation, and value flow are legible across cells, federations, and protocols, rather than stopping at the boundary of a firm or a state. The open-source contributor whose code runs in hundreds of downstream systems is attributed by all of them, not by whichever one happened to hire her. Contribution made in one community that benefits another is followed and paid.

These two properties are not features layered on top of an otherwise normal system. They are what the design is for.

What does honest mean here, exactly? The architecture does not promise attribution that is correct in the abstract — there is no such thing, and any system claiming it would be hiding the harder problem. It promises attribution that is self-correcting: many valuation methods run in parallel, scored against external signals they cannot manufacture (observed downstream impact, market comparison for similar work, member retention, calibration probes with known answers); weight routes toward the methods that predict those signals well; gamed methods lose weight automatically. The bar is not converges to truth. The bar is measurably better than the proxies current institutions route by — capital position, gatekeeping, narrative control — and improving the longer it runs. Section 04 describes the engine; what matters here is the shape of the bet: not a one-time correct answer, but a system designed to get better over decades.

Where humans can actually agree

A protocol is a designed object, and the design has a particular job: to identify the questions where agreement is actually reachable, and put only those questions in front of its members to vote on. Everything else is left to mechanisms operating inside the bounds those agreements set. People are asked to agree on what can be agreed on, rather than on grand bundles where the only honest answer is to pick a side.

Voluntary now. Voluntary always. Coexistent for as long as needed.

That is the founding principle, and the whole design has to be read in its light. The proposal is not meant to overthrow anything. It is patient, the work is generational, and what follows is the case for why that work might still be worth starting.

Why do this at all?

The attribution problem, and why current institutions cannot fix it from inside.

Every cooperation has to answer the same question: who did what, and what is it worth? That is attribution. It is the load-bearing function of any institution that pools effort and pays out value, and the honest answer is hard to find. The institutions we live inside today do not find it. They route by proxies that correlate with contribution at best and substitute for it at worst.

The proxies are familiar — who owns the slot value flows through, who sits at the gate it must pass, who holds the audience, who tells the story, what fits the institutional meter, and others in the same family. They reinforce each other: capital buys the platform that captures the attention; the gatekeeper sets the narration; the meter rewards what the gatekeeper has already chosen. The cluster is what is doing the work, even when any one mechanism looks defensible in isolation.

The Wikipedia editor who has spent fifteen years quietly rewriting biology articles so high-school students can understand them is paid nothing. Her work is read by hundreds of millions every year. More recently, it was fed into the training data of AI models that absorbed her clarity and now sell it back to those same readers. The credit, and the money, accrued elsewhere — to the platforms that surface the articles, to the AI companies whose models learned from them, to the textbook publishers who lift the prose, to whoever last said something the algorithm liked.

The volunteer who answered a Stack Overflow question in 2011 that has rescued tens of thousands of developers since, including this morning, has never been paid for it. The home cook in a small Italian town whose pasta recipe was lifted, photographed, and posted by a celebrity chef has watched it become a restaurant trend; the chef wrote the cookbook, and the cook heard about it from her grandkids on Instagram. The warehouse worker who has packed twelve hundred items per shift for six years cannot afford a one-bedroom apartment in the city the warehouse sits in; she is metered to the second on her pick rate, and the value she added accrues to the shareholders, to the brand, and to the executives whose options vest.

None of these are oversights. The shape repeats: honest enabling work — synthesis, peer support, mentorship, maintenance, mediation — fits none of the proxies cleanly and ends up systematically under-paid.

The same mechanisms also produce the rent layer. The gap fills with people whose claim runs through these proxies without the underlying contribution: passive owners, executives whose role is the gate, influencers and brand-holders, performance-review winners, meter-gamers. Each extracts a slice in exchange for performing the attribution function the underlying institution cannot.

Underneath all of them is a deeper structural failure: every one of these proxies collapses attribution to here and now. They pay the legible person in the controlling slot at the moment of payment, and they are blind in the two dimensions that honest attribution requires.

They are blind across time. The teacher's lessons matter most twenty years on, when her students' work has been priced upward by what came of it. The maintainer's quiet years are visible only when their absence would have been catastrophic. Basic research is paid out generations later, if at all. Quarterly cycles cannot see causal chains that take decades to close, and reputation that doesn't carry across generations cannot reward the people who built for them.

They are blind across space. The open-source contributor whose code runs in five hundred companies is attributed by the first one that hires her, if any. The contribution made in one community that quietly benefits another is followed by no one. Today's attribution machinery stops at the firm, the platform, the jurisdiction.

A second failure stacks on top. Even things we openly call public — clean air, immunity, working roads, basic research, a court system, a local park — are notoriously hard to fund honestly. Attribution is part of the problem (there is no slot through which a public good's value flows back to its source), but free-riding is the rest, and the institutions we built to handle it (markets, states) attribute by the same broken cluster: the loudest faction in the council chamber, the captured agency, the credentialed expert. Police, fire, and emergency response are real and necessary, but the institutions that house them are often unaccountable in exactly the ways accountability matters most.

There is a deeper reason both failures persist. Current institutions can barely tune themselves. Almost nobody is voting on the parameters that produce the widening gap between rich and poor, for instance — voting happens on candidates, on personalities, on platforms, but rarely on the actual levers. Constitutions written two centuries ago still bind the present. There is feedback in only the loosest sense: slow, indirect, easily captured, with no clean way to act on what it says.

The protocol design comes out of one question, asked against this whole landscape:

What would it look like if attribution were honest across time and space, decisions were patient, and services were accountable to the people they serve?

That is the whole bet. Not to overthrow what works or replace anything by force, but to be measurably and persistently better on the things current institutions handle worst, and to trust that over generations "measurably better" compounds into something large.

What it could look like.

A century or two out, if it actually works.

Imagine a world several generations from now in which protocols have done roughly what they were designed to do. None of what follows is a prediction; it is the direction the design points.

Work, honestly seen.

The teacher whose lessons shaped a generation of researchers is paid in her seventies, when their work has been re-priced upward and the credit traced back to her. The maintainer who kept critical infrastructure running for thirty years — while flashier projects came and went — has accumulated reward proportional to what she actually preserved. The connector who introduced the right two people in 2031, igniting work that mattered in 2055, is paid for that introduction. The mediator who quietly prevented a rupture is paid as much as the people whose collaboration she saved. The firefighter who never had a fire on her watch is paid for the years she stood ready.

The rent layer, dissolved.

The executive class — the people who currently capture spread by controlling the narrative of who did what — is gone. Not toppled. Dissolved into legitimate enabling work, openly priced. The rent niche simply has no function in a system that attributes honestly.

The rentier class is gone too. When flow is priced honestly with long horizons, most of the reasons humans accumulate stock disappear. People hold their houses, their tools, their books, their art. They do not hold portfolios of capital extracting returns from other people's labor, because the labor is paid directly. They do not transmit fungible wealth across generations, because what they have to transmit is reputation and ongoing royalty streams from their own contribution. Inheritance is honest: you pass on what you built.

Belonging, plural.

The unit of belonging is no longer a country. It is one or more protocols, each with its own bundle of values, mechanisms, mission, and culture. People belong to several at once, the way they once belonged to a country and a profession and a church simultaneously. Identity is chosen, layered, and revisable.

Local life, deliberated rather than fought.

The slow, ugly, contested decisions of neighborhood and community — heritage designations, zoning, school catchments, transit, where the new park goes, how to handle a difficult intersection — are not fought out in council chambers where the loudest faction wins and the losers absorb concentrated costs uncompensated. They are deliberated through protocol process: honest valuation of who gains and who loses, transparent compensation flowing from gainers to losers, structured input weighted by stake and verified expertise, mediation for individual hardship, long-horizon re-pricing that rewards good decisions and corrects bad ones over time. Politics, at the level most people actually experience it, becomes a craft rather than a brawl.

Public services, still here, accountable now.

The functions current states provide via organized force and presence — police, fire, emergency response, public health surveillance, infrastructure stewardship — are still provided, often better than today, but as priced presence work performed by paid participants of presence cells under constitutional authorization. Not employees of a sovereign. Not less effective for that.

States, faded.

States have not fallen — they have specialized, shrunk, and eventually faded, the way the manor and the guild faded: real things once, important for centuries, gradually displaced by something that worked better. Each generation just had slightly more attractive options than the one before, and the cumulative choices reshaped the landscape.

The human texture, unchanged.

People still fall in love, raise children, mourn their dead, make art, grow old. The protocol does not touch most of that. It just makes the scaffolding around those things less extractive, less arbitrary, less wasteful of human capacity.

How would it actually work?

The mechanism, built up slowly, in pieces that depend on each other.

The previous section sketched the destination; this one is the machinery. It is not large, but the pieces interlock, and skipping ahead makes the later parts read as more mysterious than they are.

Two primitives

The design rests on two words. Protocol is the rules: when something is allowed, when it isn't, who decides, how decisions get made. Work is everything that happens under them.

Most rules live in pricing: this contribution is worth this much, paid like this. A small, deliberately stable set of rules lives in a constitution: the questions pricing cannot answer, like when the protocol may use force, what minimum it provides to those who cannot contribute, and how it treats people outside its membership. The pricing engine is large but bounded; the constitution is small but load-bearing.

Four shapes of work

Before the rest makes sense, "work" has to widen until it fits human reality. There are four shapes, and the system pays for all of them.

• Productive work. Making something measurable. What pricing was originally designed for.
• Enabling work. Making other work possible: opportunity discovery, mentorship, gatekeeping, mediation, introduction, witnessing. Currently the most under-priced and rent-prone domain in human life.
• Presence work. Being available and ready, where the value is the readiness itself rather than a continuous stream of output. Police, fire, EMS, on-call medical, crisis response, infrastructure standby. Paid for being there in case, not for producing X.
• Constitutional work. Drafting, amending, interpreting, and deliberating about the rules pricing cannot decide. Itself paid, though the rhythm is slow on purpose.

All four are work, and all can be priced. The rent niche shrinks the moment enabling work is paid honestly, because the layers of management, brand, and credentialing that currently capture spread are mostly there to perform the attribution function the underlying institution can't.

The four-layer stack

The design implements the two primitives as a stack of four layers, each a specific move along the time and space axes the WHY section named: pricing handles time, cooperation handles space, trust carries both, and the constitution is where the meta-parameters that keep both axes honest live. From the inside out:

Pricing layer — the engine

It looks like ordinary pricing at first glance, but with a few habits ordinary markets don't have.

The deepest of these, and the one that does most of the work, is that the engine does not aim to be correct — no system honestly can be, since there is no exogenous truth about what a contribution is worth. The engine aims to be self-correcting against external signals it cannot manufacture: observed downstream impact, market comparison for similar work, member retention, adversarial probes with known answers. Pricing methods that predict those signals well gain weight; methods that drift, miscalibrate, or get gamed lose weight automatically. Which approach to valuation is right becomes empirical instead of chosen in advance. The bar is not truth. The bar is measurably better than the proxies current institutions route by — capital position, gatekeeping, narrative control — and improving the longer it runs.

Three habits implement this.

First is long-horizon re-pricing. Contributions get a best-guess price now, then re-priced over years and decades as their effects materialize. The teacher's lessons get retroactive payment when her students do important work; the maintainer's quiet years get rewarded once it becomes clear what their absence would have cost; long-horizon work like teaching, basic research, and prevention becomes fundable again. The re-pricing is anchored on signals from outside the engine — citations, dependency graphs, downstream value observed, work that continued or visibly failed without the contribution — not on a fresh internal vote. Gaming gets harder for the same reason: the verdict isn't in until decades later, and most ways to game an instantaneous score don't survive that kind of look-back against external evidence.

Second is not betting on any single valuation method. Several (peer-allocation, prediction markets, expert panels, quadratic voting, and others) run in parallel, scored against the external signals named above rather than against each other. The scoring is the load-bearing piece, and it is deliberately not internal: a method does not get credit for matching what the engine's other methods say; it gets credit for predicting what later turns out to have happened in the world. Weight routes toward methods that predict well; gamed methods lose weight automatically.

Third is honesty about overhead. Roughly 85% of value flows to contributors; about 10% pays for the protocol's own operations (the pricing engine, dispute resolution, infrastructure, constitutional work, all priced and attributed themselves); about 5% funds the safety net, basic income, long-horizon reserve, public-goods pool, and presence work. "100% to contributors" is a lie any honest system has to refuse, and 85% with transparent overhead is still dramatically better than the 30–50% labor share most current corporations deliver.

Trust layer — what pricing accumulates into

Reputation is what the pricing layer leaves behind: slow, persistent, mathematically grounded but socially textured. People build long-lived identities, pseudonymous or real-name as suits them, that carry a portable record of contribution and conduct. Reputation decays if it isn't fed by ongoing work. Some of it travels with the member everywhere; some stays partly with the cell or community where it was earned. Identity infrastructure is built to outlast individuals, with heir designations, social-recovery mechanisms, and reputation pools that survive death. A protocol that thinks in centuries needs an identity layer that does the same.

Cooperation layer — what trust enables

Pricing and trust on their own would just produce a marketplace of reputable strangers. A society needs more: shared fate, repeated interaction, slack, mutual aid, conflict resolution that preserves relationship. The cooperation layer hosts the structures markets actively dissolve.

• Cells. Bounded persistent groups of roughly five to a hundred and fifty people, sharing a revenue and reputation pool. Internally, members help each other on diffuse reciprocity — the cell is a non-pricing zone. Externally, the cell is priced as a unit.
• Federations. Cells of cells. They provide shared infrastructure (mediation, mutual-aid pools, research funding, presence-cell budgets) and let individual cells specialize.
• Mutual-aid pools. Voluntary insurance that carries members through illness, between projects, through hard patches. This is what salary actually provides as risk-pooling, made explicit.
• Mediation infrastructure. Trained, paid mediators specialized in restorative resolution. The default question is "what would let these parties continue working together," not "who's right."
• Fission as default scaling. When a protocol gets too large for its cooperation machinery, the natural move is for sub-groups to split off, form their own protocols, and federate — rather than scale internal politics to a million people.

Constitutional layer — the outer edge

The constitutional layer is short, deliberately stable, and home to every commitment pricing cannot make. The founding principle of voluntarism and exit lives here. So do the floors of provision for those who cannot contribute — children, the disabled, the very old, the broken — guaranteed regardless of whether they can be priced into anything. So do the rules about when organized force may be used and under what limits. And, perhaps most importantly, so does the protocol's posture toward parties beyond its membership: other humans, other protocols, the biosphere, future generations, entities that cannot speak for themselves.

The meta-parameters that govern the pricing engine also belong here: the bounds the engine is allowed to move inside, the mechanisms it may use to set direct parameters, the set of external signals its valuation methods are scored against, and the rules for revising all of these. Day-to-day numbers move at whatever cadence the work requires. The rules they move under are stable on purpose, because they are what members actually vote on, and voting only means anything if the thing being voted on holds still long enough to be deliberated about.

It is the smallest layer by volume and probably the most consequential by leverage. Amendments are slow on purpose. Some things should change reluctantly, after deliberation that gives weight to what's at stake.

Voting on parameters, not candidates

The system is called a protocol for a reason. Like TCP/IP or HTTP, it is the kind of thing that can be deliberately designed and revised. Members rarely vote on the day-to-day parameters themselves; those are moved by mechanisms working inside bounds. What members vote on are the meta-parameters: the bounds, the mechanisms allowed inside them, the rules for changing both. These are deliberately named in the constitution so they can be argued about. Roughly how value should be split between contributors, operations, and the safety net is a meta-parameter, while the exact split this quarter is not. What kinds of contribution count as enabling work is a meta-parameter; how a particular contribution gets priced inside that frame is the work of a mechanism. Which external signals the pricing engine scores its valuation methods against is a meta-parameter; how those methods balance each other in a given quarter is not. The response standard a presence cell is held to belongs to the constitution; the dispatch this hour does not. These are knobs on a designed object, and voting on a knob is a different kind of activity than voting on a candidate who claims they will adjust the knob.

That changes what is possible to fix. The widening gap between rich and poor in the current world has compounded for decades because there is no clean way to act on what is already known about it. Pay structures inside large firms are private; tax codes are amended by people who weren't elected to amend them specifically; wealth, once accumulated, compounds along a curve nobody is voting on. The protocol's answer is concrete. Rentier capital is removed by paying contributions directly. Land value is partly recaptured through COST or LVT. The mechanism that scores valuation methods on accuracy demotes the ones that turn out to bias upward. And the parameters themselves — slices, rates, thresholds, eligibility floors — sit where members can adjust them at the cadence the situation requires.

The floor claim here is bounding: runaway compounding patterns lose the conditions that produced them. The aspiration is stronger — that a well-tuned protocol oscillates around sustainable levels rather than trending until it breaks. Whether any particular protocol achieves the stronger version is empirical. What this design gives you, that current institutions cannot, is the apparatus to try.

Why decisions actually stick

Every voluntary system eventually has to answer this. Most never quite do. The protocol has no police; it cannot make anyone do anything. So why would a member ever honor a mediation, accept a verdict, pay a compensation transfer, or comply with a decision they would rather ignore?

The answer is connectivity.

A member of a mature protocol is connected to it through many channels at once. Their income runs through it. Their savings — held as long-vesting attribution claims rather than as capital — are anchored in it. Their reputation, which determines whether others want to work with them, lives in it. Their healthcare, through mutual-aid pools. Their dispute resolution. Their professional access. Their daily-life infrastructure. Their social embedment in cells and federations. Their access to presence services like mediation and emergency response.

When all of those route through the protocol, defying a decision costs you across all of them at once. Ignoring a mediation does not just lose the dispute — it damages your reputation across every other relationship in the protocol. Refusing a compensation transfer does not just breach one obligation — it cuts off future trade, future cell membership, future mutual aid. The protocol binds without coercing because the cost of breaking the binding is, by construction, distributed across a member's whole life within it.

This is the same enforcement mechanism premodern communities relied on before formal coercion existed: density of social, economic, and ritual connection that made non-compliance ruinous. Modern states added formal coercion on top of this density. The protocol relies on density alone — and that is sufficient, if the density is genuinely deep.

A subtle consequence: connectivity is also what makes exit meaningful. Someone who has built density across years has something real to lose by leaving — and the right to leave anyway, with reputation and personal property intact, is what keeps the density from sliding into coercion. Exit is not free of cost, but it is always available. The cost of exit is roughly the cost of disobedience, which is what makes both consequential and neither coercive.

Many protocols, one system

The end-state is not a single global protocol. One would be a monoculture, with the same capture risks, brittleness, and lack of exit that plague every monopoly. The end-state is plural: many protocols with different values, mechanisms, missions, and cultures, with people belonging to several at once.

Some protocols specialize by domain: the ones built for high-velocity creative work look quite different from the ones built for slow craftsmanship or contemplative life. Others specialize by place. When membership clusters in a neighborhood or city, members naturally form (or fork into) a place-aware protocol that handles what local government currently handles: zoning, building codes, public health, neighborhood services. And as a non-negotiable commitment of the wider system, there is always at least one open-membership protocol per region, with low friction to join and baseline floors guaranteed, so that "non-protocol human" is a temporary state rather than a permanent disenfranchisement.

Above all of this sits a thin inter-protocol layer: shared standards for identity portability and reputation legibility (so a member in good standing of one protocol can be recognized by another); bridges for moving value between protocols; mediation paths for inter-protocol disputes; a universal-personhood commitment binding every participating protocol to treat all humans, members or not, as persons. Where existential risks exceed any single protocol's scope — pandemics, climate stabilization, asteroid defense — a specialized federated body monitors and responds, funded by a small levy on all participating protocols, with authority delegated specifically and only for that purpose.

When a protocol goes badly wrong — turns into a cult, a scam, or a coercive system — the response is non-coercive but real. Verification audits remove its standing in the inter-protocol layer. Reputation pressure costs it access to trade and recruits. In the extreme case, the rest of the system collectively ceases all exchange with it, and members trying to leave are received by legitimate protocols with low-trust onboarding. This is roughly how the international system handles pariah states today, and the protocol system can probably do better at it, because exit is built in at the foundation and members hemorrhage out faster than they do from bad states.

How we get there from here.

Patient, legal, voluntary, and mostly the work of generations not yet born.

The previous section was the destination. None of it describes what the first protocols look like. The path matters as much as the design, and its discipline is the hardest thing about the project: it asks people to work for a future they will not see.

Coexist with everything that already exists.

For the entire build phase — most of the protocol's existence, and the only phase any of us alive today will witness — the protocol operates as a law-abiding voluntary association. This is a constitutional commitment, not a tactic.

During its coexistence phase, the protocol operates entirely within the legal frameworks of the jurisdictions where its members reside. It does not challenge state sovereignty, evade state authority, or position itself as an alternative state. Members comply with all applicable laws.

In practice: pay taxes, register entities, sign contracts, respect intellectual property, comply with employment law where it applies, use existing courts for disputes legal systems have jurisdiction over. The protocol does what existing institutions are bad at, and leaves what they are good at — deep contract enforcement, currency, borders, infrastructure they already paid for — to them.

Picture a software developer today. She is a citizen of a country, a resident of a state, an employee of a corporation, a member of a labor union, a contributor to an open-source foundation, a Catholic, a LinkedIn user. All of these coexist. Adding membership in a protocol — earning priced attribution for her contributions, voting in her local protocol-mediated heritage process, contributing to a presence cell for neighborhood mediation — requires nothing to be torn down. It just adds another layer of voluntary affiliation.

Adoption is patient embedding, not vampire-attack.

The protocol does not try to win by persuading the world of its civilizational vision. The vision motivates the people inside; what's offered to anyone outside is utility, on a few specific things, done well. A handful of working principles follow:

• Stay reversible. Entry is low-commitment by design: no need to give up your job, bank account, or passport. Route some work through and see if it serves you better.
• Build boring competence on a few specific things. The pitch becomes "the best place to get paid fairly for open-source contributions," or "the most credible neutral process for the next council fight," or "the most accountable way to fund the local fire cell." Not "the new civilization."
• Diversify infrastructure, especially payment rails. Most movements of this shape die quietly of de-banking, not of raids.
• Cultivate allies inside the state: people in tax compliance, employment data, anti-fraud, willing council members. There is always a faction inside any state that benefits from the protocol existing, and it pays to know who they are.

What to honestly tell yourself.

If you find yourself drawn to this work, the discipline is real and unforgiving. The first protocols will be small, awkward, and prone to failure; most of the people who build them well will not be remembered; the destination, if it arrives, will arrive in someone else's lifetime. The trade is your best work for a future that will never thank you.

What we control is the present discipline: build well, stay legal, accumulate trust, build connectivity deliberately, respect voluntarism absolutely, honor commitments toward those who never consented. And trust that "measurably better" sustained over centuries is enough.

What this asks of you, in the short term.

Almost nothing dramatic. Notice, in your own life, the work that's currently invisible, the local decisions that go badly because the machinery is wrong for them, the presence work nobody is paying for. And if a protocol-shaped initiative crosses your path (an attribution experiment in some open-source project, a retrospective public-goods funding round, a local council willing to use deliberation infrastructure as input, a neighborhood mediation cell looking for funders), give it some real attention and see how it does.

If any of it rings true, the most useful thing is to talk about it with people you trust. Designs of this kind don't fail technically. They fail by being forgotten between the generation that cares about them and the next.

Voluntary now. Voluntary always. Coexistent for as long as needed.