Propositions

There is something slightly misleading about the way modern economies speak about growth.

For most of human history, energy arrived as a steady flow. What could be used this year looked much like what would be available next year. Planning, scarcity, and effort all took place within that rhythm.

Industrial society changed this without ever fully renaming it. Large parts of what now counts as ordinary economic activity depend on releasing energy that was accumulated over vast stretches of time and then spent very quickly.

The language of income still works. Budgets balance. Productivity graphs rise and fall. But the meaning of those numbers has quietly shifted.

Once you notice this, many familiar arguments begin to feel oddly misaligned.

Economic activity is often described in terms of how much is happening: how much energy is used, how much material is processed, how many hours are worked.

This focus made sense when energy was cheap and abundant enough that its use rarely constrained outcomes. Under those conditions, doing more often coincided with achieving more.

As those conditions change, something becomes clearer. Energy and materials do not disappear when they are used. They are transformed, dissipated, and made unavailable for other purposes.

What increases is not value by default, but expenditure.

Living systems persist by making the future easier to inhabit.

A forest that stabilises soil, a skill that shortens learning time, a language that enables coordination—none of these consume much energy once established, yet all make many future actions possible that were previously difficult or impossible.

Some arrangements do the opposite. They produce short-term gains while narrowing what can be done next, often by creating dependencies, fragility, or irreversible damage.

If value exists anywhere beyond immediate preference, it appears here.

For a long time, work solved two problems at once. It produced what people needed, and it determined who was allowed access to those needs.

As long as those functions aligned, tying survival to labour appeared both practical and fair.

That alignment is loosening. Many necessities can now be produced with little human labour, while many activities that matter most to long-term stability—care, learning, coordination—do not resemble jobs in any conventional sense.

Under these conditions, societies often respond by inventing activity whose main purpose is to justify entitlement rather than to improve outcomes.

Any economy that intends to persist must decide how people remain part of it.

Guaranteeing participation does not mean promising comfort, abundance, or uniform outcomes. It means ensuring that exclusion is not used as the primary means of coordination.

When participation is bounded by physical reality rather than enforced scarcity, contribution can be aligned with what actually improves future conditions rather than with what merely satisfies accounting proxies.

Scarcity remains. Trade-offs remain. What changes is how they are handled.

Societies remember in more ways than one.

Some memory lives in people: skills, habits, expectations. Some lives in things: buildings, tools, institutions, laws. Much of it is not written down at all, but embedded in how easily certain actions can be taken and how difficult others become.

When these structures are well aligned with reality, they make sensible behaviour easy. When they are misaligned, effort increases while outcomes worsen.

An economy is not just a system of exchange. It is a system of memory.

Complex systems feel powerful when they are first built.

They coordinate large populations, move resources efficiently, and solve problems that once seemed impossible. Over time, however, something quieter happens. Keeping the system running begins to require more effort, more oversight, more energy.

When the surrounding conditions are generous, this cost stays hidden. When they tighten, complexity starts to weigh on the system that produced it.

Complexity is never free. It borrows from the future.

Care is often spoken about in emotional terms, as something personal or moral.

But care also has a structural dimension. It preserves capacity. It prevents damage from becoming irreversible. It keeps options open that would otherwise close quietly and permanently.

When care is absent, systems do not fail immediately. They degrade. Possibilities narrow. Recovery becomes expensive or impossible.

Seen this way, care is not a luxury. It is a form of configuration maintenance.

Many of the most important economic actions leave little trace.

A problem that never materialises does not show up in statistics. A crisis that is quietly avoided rarely becomes a headline. What remains visible is the effort that follows failure, not the attention that prevented it.

This creates a bias. Systems that reward visible response over quiet prevention slowly train themselves to fail first and fix later.

Over time, this is expensive.

Stability is often mistaken for stillness.

In living systems, stability comes from continual adjustment. Forests shift, bodies adapt, languages evolve. What remains stable is not the arrangement, but the ability to respond without breaking.

Economic systems that confuse stability with freezing existing structures become brittle. When conditions change, they crack rather than flex.

A stable economy is one that keeps moving—carefully.

Economic decisions are often discussed as if time were neutral.

Costs and benefits are discounted, averaged, or projected forward as though losses and gains could be freely exchanged across decades. Yet many changes do not reverse. Forests regrow slowly, if at all. Species vanish permanently. Skills atrophy. Trust erodes quietly.

Time does not treat all actions equally. Some close doors forever.

Many transactions look clean at the moment they occur.

Payment is made. Goods change hands. Accounts balance. What does not appear is what has been displaced: the wear transferred elsewhere, the waste deferred, the future constrained.

Costs often arrive later, or in places no longer connected to the original decision.

Scarcity is usually spoken of as a fixed condition.

In practice, it shifts with technology, organisation, and configuration. What is scarce in one arrangement becomes abundant in another. What seems plentiful can vanish when coordination fails.

Scarcity is not simply "how much exists". It is how easily something can be accessed, reused, or substituted.

Technological optimism often rests on substitution.

If one resource runs short, another will replace it. If one process becomes expensive, a new one will emerge. This has sometimes been true — but not without friction, delay, or loss.

Some substitutions are slow. Some are partial. Some trade one constraint for another.

When total output increases, tensions often soften.

Losses in one place are offset by gains elsewhere. Conflict can be deferred. Inequality feels tolerable when absolute conditions improve.

When growth slows, these buffers vanish. Distributional questions move from background noise to central concern.

Money coordinates action by compressing complexity.

Prices summarise many variables into a single figure. This is powerful, but also dangerous. When the signal detaches from physical reality, coordination continues even as the underlying system degrades.

The signal remains legible long after its referent has changed.

Some societies achieve more with less.

They waste less energy resolving conflict, duplicating effort, or repairing avoidable damage. Much of their "wealth" lies not in assets, but in shared understanding and trust.

Coordination reduces the cost of acting together.

Systems prefer what can be counted.

Simple metrics travel well. They fit dashboards, headlines, incentives. Unfortunately, the most important things often resist simplification.

As legibility increases, nuance is lost. What cannot be seen stops being protected.

Stable systems tolerate inefficiency.

Unstable ones do not. During transitions, buffers shrink. Old supports weaken before new ones fully form. Errors that were once survivable become critical.

This is when accounting honesty matters most.

No economy escapes physical reality.

What differs is whether limits are acknowledged early or imposed later. Ignoring constraints does not remove them. It merely postpones their effects until they arrive uninvited.

Denial is not neutral. It is a decision with predictable consequences.

There is a tempting shortcut in discussions of configuration.

If value lies in arrangement rather than material, perhaps everything can be digitised. Data is cheap to copy. Symbols travel freely. The constraints that bind physical systems might be escaped through abstraction.

This is a misunderstanding. Configuration includes bodies, skills, relationships, institutions, infrastructures, and ecological arrangements. Most of it cannot be uploaded. All of it must be maintained.

A forest is not a database of trees. A skill is not a video of someone performing it. A functioning institution is not its org chart.

Configuration that ignores its physical substrate is not configuration at all. It is description.

If value is configurations that expand future option space, a practical question follows: how is option space measured?

The question splits in two.

*Global* option space — a scalar over a configuration capturing how much future possibility a state contains — remains genuinely open. Backward-looking measures exist: Cronin's Assembly Theory counts the minimum steps required to construct an object; Bennett's logical depth captures computational history embedded in an arrangement. Both measure history, not forward potential. Candidate forward-looking formalisations exist — reachable configurations under a bounded exergy budget, Kauffman's adjacent possible, state-space cones in control theory — but none has been shown to be economically operational at the state level.

*Local* option space — the differential effect of a move on the reachable set — is tractable. Δω evaluation requires only directional comparison between pre-move and post-move reachable sets, not a global scalar. Mass extinction, monoculture conversion, fossil-fuel lock-in to 4°C are unambiguously option-space-degrading even where global ω cannot be computed. See option-space-as-chess-moves.

One family of candidates is already ruled out, and the razor-blade case is what rules it out: letting a child play with razor blades widens the immediate option set and collapses the trajectory's. The instrument that approves it is local in scope (counting only what is reachable within a radius), local in time (counting what is executable at this instant — a reading that does not merely lose precision but can carry the opposite sign, since the frontier stays populated long after the trajectory has closed), and local in term (counting for one entity rather than for the agent, the non-agent, and the relationship between them). The failure is detectable without solving the measurement problem: read the trajectory-long sign triple across those three terms, and a mixed sign — any + alongside any − — marks a verdict that survives only under a flattering aggregation. In the two-agent reading, SUM rising while MIN falls is the same signal. A move that needs a favourable lens has failed. This narrows the candidate space; it does not close it.

What remains open is the global state measure. Until it lands, "option space" at the state level functions as a direction for measurement rather than a measurement. At the move level, the framework already ships with a usable decision procedure.

If throughput cannot serve as the measure of success, what replaces it?

The instinctive move is to nominate a new quantity to maximise — durable flourishing, well-being, sustainability, configuration quality. Each of these names something real, but each runs into the same trouble: maximising a single scalar across many living systems requires an aggregation rule, and any aggregation rule trades off against some class of life it weighs lightly.

A different shape is available. Instead of asking *which configuration is best*, ask *which moves preserve the future*. The objective becomes a no-regret criterion over moves rather than an optimisation over states.

Prefer changes that keep the reachable set of life-supporting configurations open for all major classes of life across all relevant horizons. Reject changes that asymptotically close that reachable set for any such class at any such horizon.

This is closer to medicine's *first do no harm* than to economics' *maximise utility*. It tolerates incomparability between dominant configurations — it does not need to pick a winner among non-dominated states. It only needs to refuse moves that close categorical possibility.

Under this framing, the economy does not grow or shrink as a whole. It reorients toward moves that keep major classes of life in play.

If value lies in configurations that expand future option space, a question follows immediately: option space *for whom*?

The same physical state contains very different reachable sets depending on which observer asks. A configuration that opens many futures for a corporation may foreclose them for a watershed; a configuration that secures a phylum's continuation may eliminate a species within it.

Standard responses are unsatisfactory. Aggregating across all conceivable observers requires a weighting rule that simply pushes the disagreement up one level. Choosing humanity as the privileged observer is parochial and unstable across timescales.

A different move is available, and physics has already made it. Stephen Wolfram's *Observers Like Us* argues that the laws of physics we observe aren't intrinsic to the underlying structure of reality — they're the slice that observers of our class (computationally bounded, persistent, sequentialising) access. Different observer-classes carve up the same underlying structure differently. The principled move is to make the observer-class explicit, not to deny it.

The parallel for value accounting: option space isn't intrinsic to configuration space. It's the slice that **life-persisting observers** access — the set of futures in which living entities continue to exist and make distinctions. Reachable configurations are tagged with the observer-class that can reach them.

Formally: R_living(C, B, T) — configurations reachable from C, given an exergy budget B and horizon T, in which living entities continue to exist and make distinctions. Observer-relative by design, not by concession.

This gives the "for whom" question a structural answer. Not utilitarian aggregation across individuals — most living individuals will be replaced over any meaningful horizon. The criterion is at the **category** level: preserves the capacity for life to exist as a category, with current functional diversity, over the relevant horizon.

When this work speaks of economies "spending energy" or "running on energy income," the physically precise term is exergy: the portion of energy available to do useful work.

Energy itself is conserved. It cannot be spent. What is spent is its ability to do work — its distance from thermodynamic equilibrium with the surroundings. A battery contains the same energy before and after it powers a laptop; what changed is that a usable gradient has become dispersed heat.

This distinction is implicit throughout. Where the essay uses "energy" in economic contexts, "exergy" is what the physics supports.

In accessible prose, "energy" reads better and is familiar. In formal claims, "exergy" is precise.

Earth receives solar exergy at roughly 175 PW continuously. Human civilisation dissipates roughly 18 TW. By the raw thermodynamic budget, we are nowhere near a ceiling.

This matters because a common reading of "bounded energy flux" is that economic activity is hitting the second law. At planetary scale, it is not. The active constraints are elsewhere.

Waste heat and waste configurations must go somewhere, and the biosphere, oceans, and atmosphere are the sink. Specific materials — phosphorus, rare earths, fresh water, topsoil — are bounded not by total quantity but by extraction and regeneration rates. Biospheric integrity governs which configurations remain stable at all.

"Bounded energy flux" in this work names these practical bindings, not the distant thermodynamic ceiling. The ceiling is far; the floor of what the biosphere can absorb is close.

If global option space is hard to measure, the practical question is whether it can be evaluated locally — move by move — well enough to act on.

Chess engines do not compute the full game tree. They evaluate moves by whether the resulting position has more or fewer good options. Did this move open up the position or close it? Even directionally, this is enough to play.

The same operation is available here. Every policy decision, infrastructure commitment, land-use change, institutional reform is a move on the configuration. The question is local: does this move expand or contract the reachable set of life-supporting configurations, over the horizon a class of life depends on?

Mass extinction events, monoculture conversions, fossil-fuel commitments locking in 4°C are unambiguously option-space-degrading even when global option space after the move is non-computable. Local gradient evaluation is tractable where global value computation isn't.

The locality is in the operator, not in the object. Δ is taken one move at a time; ω is not read at an instant, inside a radius, or for a single observer. Every evaluation is scoped to a stated boundary B and a stated horizon T — the reachable set of life-supporting configurations, for a named class of observers, over the horizon that class depends on. Read the object locally instead of the operator and you have a different measure, one that can carry the opposite sign: this is the razor-blade case, and it is why B and T are stated rather than assumed.

This is not new. Dave Snowden has articulated next best move as the central practical primitive in complex domains since Cynefin took its modern form. Aubin's viability theory formalises admissible trajectories under non-empty constraint sets. This work inherits both lineages and adds physical grounding — moves are evaluated over an explicit exergy budget and horizon, restricted to the class of observers that maintain themselves and persist.

Labour has long been described as the source of economic value — the thing that turns raw nature into wealth, the substance that prices are supposed to track.

A different reading is available. For most of human history, labour was the mechanism by which people were allocated access to life-supporting configurations under conditions of scarcity. It was the coordination machinery scarcity required, not the thing being coordinated.

The two roles were rarely distinguished because, under those conditions, they coincided closely enough. Where production needed many hands, allocating access by hours worked was a serviceable approximation. The allocation mechanism tracked what mattered — roughly, locally, well enough to organise life.

As the composition of scarcity shifts — energy carriers, materials, automation, machine cognition — the mechanism begins to drift from what it was tracking. Hours worked no longer index contribution to configuration, and never did directly; they indexed eligibility under a particular scarcity regime.

When the regime changes, the index does not automatically re-anchor. It keeps measuring eligibility under a regime that no longer obtains.

Option-space change is not symmetric in time or in scale.

A configuration that took millennia to assemble — a soil profile, a language community, a coral reef, a trust-bearing institution — can be dismantled in years or hours. The reverse rarely holds. The arrangements that expand future possibility are typically built slowly, through processes that integrate many small adjustments across many parts of a system. The arrangements that contract it are typically broken quickly, through processes that need only one well-placed failure.

This is not a moral observation. It is a structural one. Build-up processes are systemic and gradient-following; breakdown processes are local and gradient-releasing. The two have different characteristic rates because they are doing different physical work.

Where this asymmetry holds, expected-value reasoning misprices irreversible loss. A fast-local certain destruction cannot be reliably offset by a slow-systemic uncertain expansion. The no-regret stance is not conservatism by temperament; it is the operationally tighter rule under the asymmetry.

Each configured arrangement makes further configurations reachable that were not reachable before.

An enzyme makes a reaction accessible that chemistry alone did not provide. A notation makes a proof accessible that the same idea, unwritten, did not. A road makes a settlement accessible that geography alone did not. A standard makes an interoperation accessible that two compatible systems, unstandardised, did not.

Once enough configuration has accumulated, the rate at which new configurations become reachable grows with what is already there. This is not a moral observation. It is a structural one — the same mechanism that makes destruction fast and local, run in reverse, makes expansion slow but compounding. Each addition opens a neighbourhood of further additions that the prior state did not contain.

The compounding is not unlimited and it is not always benign. Configurations that expand option space for one observer can contract it for another. Substrates that took long build-up can be removed faster than they accrued. The asymmetry of rates between expansion and destruction is what makes the slow generative process structurally vulnerable to the fast destructive one.

Where this generativity holds, value is not stored in configurations as a stock; it is produced by configurations as a flow into the adjacent possible.

An economy has to do two things at once: decide who gets access to the configurations that support life, and aim collective attention at the work worth doing.

For most of history a single mechanism did both. Labour rationed access — you ate if you worked — and pointed attention, because effort flowed to where it was paid. The two jobs rode on one rail, and while they rode together it was natural to mistake the rail for the thing it carried.

That rail is coming loose. As automation decouples production from labour and the composition of scarcity shifts from things-made to sinks-and-substrate, the mechanism that did both jobs begins to fail at both. The instrument reached for instead — the market price — was only ever doing one of them, and only on one board.

The reflex is to ask how to price the missing costs better. But a better price is still a price: one number, on one board, drifting from what it tracks. The deeper question is what kind of problem coordination is in the first place.

A different shape is available. A market asks for one thing — the price at which the board clears. A game of chess asks something else of every player: is this move better or worse than that one, given the position I can see? No player computes the whole tree; none needs a global valuation; directional, local evaluation is enough to play well. If value is generated move by move, coordinating it may be less like finding a price and more like playing well across many boards at once.

Ask what an economy is *for*, and the usual answers name a finish line: a target rate of growth, a level of output, a steady state to be reached and held. Each treats the economy as a game to be won — played to arrive somewhere and stop.

There is an older distinction that cuts the other way. A finite game is played to win; it ends when someone does. An infinite game is played to continue the play. Its purpose is not to reach a final state but to keep the game going, and to bring others in. The two are not different strategies for the same game. They are different kinds of game.

Read the viable objective again in that light. To prefer moves that keep the reachable set of life-supporting futures open, and to refuse moves that asymptotically close it, is not a finite player's strategy for winning. It is the rule of an infinite game: make no move that ends the play. The point was never to win the economy. The point is to keep it open, and to keep the distinctions coming.

Stated in the terms of this work, option space *is* the infinite game given a physical substrate. R_living is the set of futures in which living systems continue to exist and make distinctions — which is to say, continue to play. To preserve it is to keep the game in play; to empty it is to end the game for some class of player, permanently.

Naming it does real work. It tells you what kind of thing the objective is, so you stop looking for the win that was never there — and it restores the register the diagnostic frame loses: an infinite game is an invitation, not an emergency. The work has said as much from the outside in — an invitation to an infinite game, the point of which is not to win or finish it but to keep it going and bring others in, continuing the work begun by people like Buckminster Fuller. This proposition brings that framing inside. Fuller gives the means — ever more with ever less, in service of all life — and Carse gives the purpose those means serve.

This one is still taking shape, and it should be read that way.

The work already says value lives in what is reachable, not in what is stored. There seems to be a corollary that has not been stated: value that is held but is not reaching anything is, in option-space terms, closer to dead than to stored. It sits on the books as wealth while doing little for the reachable set.

Two things that look unrelated may be the same thing. Money hoarded — kept precisely so it is not exchanged — converts the live optionality money is good for into a static claim. A privately owned car stores a great deal of material value in one configuration that is idle for most of its life. On paper both are wealth. In option-space terms both are capital frozen out of reach for as long as they stay that way.

If that holds, the demurrage intuition — money that decays if hoarded — was never really about money or about fairness. It would be one case of a more general thing: configurations that lock value out of reach are mispriced as wealth.

What stops this being glib is that not all holding is dead. A seed bank, a reserve, savings against a shock, slack in a system — these are held and not reaching, and they are plainly valuable, because they widen the reachable set against future contingency. So the claim is not "storage is waste." The real work, still undone, is drawing the line cleanly between value held as live contingency and value frozen with no contingent return.

The sharpest objection, shipped with this claim: that thawing a frozen asset only moves option space around rather than making more of it — a shared car reaches more people but is the same car, so this is redistribution wearing physics vocabulary. The honest answer is that the distinction is real (a thaw can lower the frozen capital a given set of journeys requires, which is a genuine reduction and not a transfer) but unproven: no case here measures a thaw that demonstrably expands the reachable set rather than merely transferring it. Until one does, the claim holds only for the cleanest cases — redundant duplicates, hoards held as hoards — and that is exactly why it sits at forming.

This entered as forming and has since climbed: the coordination/enclosure edge has held across several worked applications, and the work now leans on it.

The work has mostly drawn the edge of the reachable set with physics: what an exergy budget, a material stock, or a biospheric sink will allow. There seems to be a second edge, drawn by something else. Some configurations are physically buildable right now — every component exists, every step is known — and they still cannot be reached. Not because nature forbids them, but because the pieces are held apart by who owns what, who is paid for what, and which institutions will or will not combine.

Three cases point the same way. The best possible phone is not buildable, because the strongest pieces sit inside Apple, Google, Samsung, Qualcomm and others who will not pool the rights to assemble them into one device. A cure for a disease that affects a few hundred people goes unmade, not because the biology is harder, but because the configuration is unprofitable to assemble. And the sharp case: a billionaire with that disease cannot buy the cure, because no amount of willingness to pay reaches a configuration that better coordination between hospitals, labs, and researchers would have produced but did not. The money is real and the demand is total; the assembled know-how simply does not exist to be bought.

That last case is the one that does the work. It says the boundary is not always scarcity of means. Unlimited means can sit on one side of a configuration that is physically permitted and still out of reach — held there by coordination and enclosure, not by physics. If that holds, market efficiency fails from the inside: the price system cannot clear a trade for a thing that has never been assembled, however much someone would pay.

What stops this being a complaint about monopoly is that the same enclosures often exist to fund the assembling. Patents and profits are how the know-how got created in the first place. So the claim is not "enclosure is theft." The undone work is drawing the line between enclosure that funds the building of capability and enclosure that only freezes capability already built.

The sharpest objection, shipped with this claim: that the unreachable cure is not news but a textbook fixed-cost problem — markets are already known to under-provide any good whose assembly cost exceeds a single buyer's willingness to pay times the number of people it would serve. The honest answer is that this may be a re-description of the public-goods account in configuration vocabulary rather than an addition to it; whether the coordination edge names something the fixed-cost story misses is the live frontier. What the configuration framing adds — and what has held across the worked cases — is the diagnosis that the boundary is reachability, not means, and the prediction that pooling the held pieces makes physically-buildable configurations reachable with no change in the physical budget. That the boundary has a coordination edge at all is now load-bearing; whether that edge is strictly more than the textbook public-goods case is the part still being settled.

This is still forming, and it is the load-bearing one — so read it with the most care and the least credence.

Start from the question of what actually bounds the reachable. The work has already moved the wall off energy: Earth receives solar exergy at ~175 PW and civilisation dissipates ~18 TW, so the thermodynamic ceiling is distant (binding-constraint). Atoms are abundant too. What is scarce is knowing which arrangement of available atoms would expand the reachable set, and being able to assemble and coordinate it.

Distinction Physics gives this a precise spine. Every change of configuration is a trajectory through a space of distinctions, and its cost is the action S = ∫E dt — energy integrated over time, the same quantity classical mechanics calls action (DP, Module 4). A desired configuration is reachable when some feasible trajectory of bounded action reaches it. So the cost of reaching anything is action — energy × time — and "this problem is hard" turns out to mean only "this trajectory requires a great deal of action." Hardness is a magnitude, not a category. Nothing a configuration's difficulty alone can do puts it outside the possibility space.

What does put a configuration outside it is a physical invariant — a wall physics forbids crossing at any action (a conservation law, the entropy floor, an irreversibility on the relevant horizon). Those are the only true walls, and binding-constraint already names them. Everything else we call hard, scarce, unprofitable, or unreachable is a question of how much action a goal needs, whether anyone can find a low-action path to it, and whether anyone mounts the action at all.

Here is where intelligence enters, in DP's own terms. In physics, nature selects the least-action path for free — δS = 0 is a law of motion. But among configurations that do not yet exist, nothing selects it automatically; the low-action path has to be searched for. That search is what pattern-intelligence is: the faculty that approximates the least-action trajectory to a desired configuration — doing deliberately what nature does for nothing. It does not repeal the action cost; it finds the short way through, so the same goal costs far less action than blind search would.

So the framing, offered as direction rather than a settled law: the possibility space yields to action × pattern-intelligence. Action (energy × time) is the fuel; pattern-intelligence is what spends it on the right path instead of wandering — DP's "excess action" is exactly the wandering intelligence removes. Read this way, a step-change in pattern-intelligence — machine pattern-intelligence at scale — is potentially the largest lever we have ever had on the reachable set: tools that improve thinking that build the next tools, each round finding lower-action paths the last could not (configuration-generates-configuration, run forward and faster).

Two edges keep this honest, and they belong inside the claim, not after it.

First, it relaxes the search, assembly, and coordination constraints — not the physical ones. The sinks, the specific materials, the biosphere, entropy itself stay exactly where binding-constraint put them, and pattern-intelligence aimed at throughput presses harder against them, not softer. The lever is real; the floor is unchanged.

Second, the asymmetry cuts both ways. A bigger lever moves the world faster in whichever direction it points, and option-space destruction is fast and local while expansion is slow and systemic (asymmetry-of-option-space-change). Pattern-intelligence aimed at a configuration that is itself shrinking the reachable set accelerates foreclosure. By default — unaimed, or aimed at yield under a price that cannot see the sink — that is the cheaper direction. So this is better called the greatest available amplifier than the greatest lever: it makes the question of aim (coordination-as-move-evaluation) and of objective (viable-objective) more urgent, not less. In the terms of infinite-game, it is a far more powerful way to either keep the game open or end it.

That is also why money is not the thing. Money is neither action nor pattern-intelligence; it is a claim on them, and it converts to mounted action only where coordination and incentive carry it there. For a configuration that is unprofitable to assemble or whose pieces are held apart, the claim never converts — which is why unlimited willingness-to-pay can sit beside a physically-permitted cure whose low-action path no one ever traverses. The path is not forbidden and may not even be especially long; it is simply never taken.

The sharpest objection, shipped with this claim: that every general-purpose technology was, in its moment, going to end scarcity — steam, electricity, the Green Revolution, nuclear, the internet — and each relaxed a specific constraint only for the constraint to relocate, often to an irreversibility (aquifer drawdown, nuclear waste, attention capture, data-centre load). The base rate for "this one finally escapes scarcity" is poor. And the lever has a footprint: under the prevailing aim — yield under a price that cannot see the sink — a cheaper way to produce is first of all more production, so the rebound loads the binding constraints before it relaxes anything. The honest answer concedes both, which is why the claim is narrowed to relaxing the non-physical, arrangement-bound constraints that have actually been binding, while the physical floor stands and is pushed harder by default. That pattern-intelligence acts on the configuration-finding process itself may make it different in kind — but that earns a wider error bar, not a presumption of rescue. Hence forming, and hence amplifier rather than lever.

This one is still forming, and it is worth saying where it came from: it generalised out of a single worked case (a north-Wales valley evaluated against this framework), which is how the work is meant to grow — an application throwing off a principle.

The case kept hitting a wall, and the wall was not physics. The energy was there, the demand was there, the configuration was buildable. What stopped it was the regulated grid: the connection queue, the licensing boundary that turns a shared network into a regulated utility, the powers held far above the locality. That is coordination-bounds-reachability wearing infrastructure: a configuration fenced not by atoms but by who controls the wire and the rules.

The general shape: the binding boundary on a reachable configuration is often a control chokepoint — a centralised point where access is permissioned — not a physical limit. And a chokepoint has a double nature worth naming. The grid is the one place where electrical power and political power are the same object: centralised generation feeds centralised wires feeds centralised permission feeds centralised control. So distributing the physical thing distributes the political thing with it. To route around the chokepoint is not only an engineering move; it is a redistribution of control.

The move the case pointed at: architect at the largest scale that still stays below the boundary, so the centralised system becomes optional — a buffer, not the spine. For energy that means generating, storing and using behind the meter, at building and cluster scale, where no queue and no licence reach.

Two things keep this honest and belong inside the claim. First, you buy autonomy with redundancy. The reason shared grids exist is diversity: separate peaks need less total storage than the sum of self-sufficient units. Drop below the boundary and each unit over-provisions for its own worst case. So the rule is the largest unit that stays below the boundary, not the smallest possible one — a privately-owned cluster recovers most of the pooling a public grid would have given, without crossing into it. Second, the boundary re-enters at the point of exchange: the moment you sell the thing across the boundary to others, you can trip the regulation you routed around. The value-capture has to be built so it is not, in the regulated sense, supply.

And the gate that decides whether this brings greater life to all or only to some: it has to be owned collectively. Routed around individually, decentralisation is just the comfortable seceding to private abundance, leaving everyone who cannot afford the kit stranded on a thinning shared system whose fixed costs they now carry alone. Owned together, the same hardware becomes greater life to all. The hardware decentralises; the ownership must not.

The sharpest objection, shipped with this claim: that routing around shared infrastructure is, historically, exactly how the comfortable secede — gated communities, private security, private schools — and calling it collective only fixes it inside the cell, not between cells. A rich valley can island itself; a poor one cannot, and a county of self-sufficient wealthy cells beside stranded poor ones is not greater life to all, it is the death spiral moved up a level. The honest answer is that collective ownership passes the gate within a cell but does not by itself supply the inter-cell redistribution a centralised, universal-service system was also doing; the principle routes around the control chokepoint without yet saying what replaces the solidarity the chokepoint also carried. A second edge: regulatory boundaries move — push enough value below a boundary and it tends to follow you down to recapture it — so routing-around may work at small scale and self-defeat at scale. Both are why this sits at forming: it names a real and powerful move and does not yet close the equity and scale questions it opens.

This one is still forming, and it generalised out of a worked evaluation — a six-hat run on whether a UK law firm could take the Saudi football money on these principles — which is how the work grows: an application throwing off a principle.

The setup is general. A large flow of value comes from a source tied to credible harm, and you are tempted to take some of it ethically — to do the clean version, raise a floor, build a good institution, and tell yourself the good work justifies the proximity to the money.

Here is the counterintuitive core. When the money is buying legitimacy, your legitimacy is the product. The credibility you build doing the clean work does not stay attached to the clean work; your name is fungible, and that name then sits on the thing that harms. So sincerity can launder more efficiently than cynicism: a visibly principled actor is a better reputational asset for the harm than a cynical one, precisely because more credible. The cleaner you are, the better the cover you supply.

It follows that cleanliness and scale are inversely correlated. The large money is indivisible from the laundering, because presence is indivisible — the seat at the lucrative table is bought by doing the thing that harms. A genuinely clean configuration does exist, but it is small and adversarial: you make your money by resisting the harm rather than receiving its coin — acting for the harmed party, for the financier who can impose a floor, for the union, in a forum the harm-doer cannot police. You profit by swimming across the current, not riding it. And the tell that you have crossed back over is the word significant: the moment a clean role has to be significant to be worth it, it is reaching for the harm-doer's shilling.

Two things keep this honest. First, less to none is a gate, not a sum: a real benefit to some does not net off the cover you manufacture for harm to many, so you cannot offset one against the other. Second, even the adversarial clean position is rarely perfectly clean — you may have reached it on credibility the wider market gave you — so it is floor-raise with a residual cover cost, a difference of degree, not a halo.

The operational line, the part worth carrying to any loaded domain: four tests separate floor-raising from laundering. Counterparty — is your client the harmed party, or a non-sovereign actor imposing the floor, and never the harm-doer. Forum — does the remedy attach where the harm-doer cannot police it. Brand-spend — does your name, when spoken, subtract legitimacy from the harm rather than supply it. Indivisibility — because presence is indivisible, it is a fork, not a portfolio you can balance: you cannot serve both sides and call the good half your defence.

The sharpest objection, shipped with this claim: that it perversely tells good actors to abandon the field to bad ones — if sincerity launders more, the ethical actor should withdraw and leave the harm-doer to the cynic, which makes the world worse, not better; and everyone, even the worst client, is entitled to representation. The honest answer is that the claim does not say withdraw from the domain — it says move to the other side of it (act for the harmed, the financier, the union), which keeps a principled actor engaged exactly where they raise a floor; the conflation it refuses is between leaving the harm-doer and leaving the field. But the concession is real: in some domains there is no adversarial perch, and there the principle does counsel exit — and whether that helps, or merely leaves the cynics unopposed, is genuinely unresolved. That, and the fact that the line between subtracting and supplying legitimacy is a judgment rather than a measurement, is why this sits at forming.

This one is still forming, and it came out of a gate being used in anger rather than out of theory. The working notes carry a design gate called the bridge test: a proposed configuration fails if reaching it requires a campaign, a founding meeting, or a heroic act of collective will — every step must be individually rational for its adopter on day-one terms. The test went into its first adversarial run as a binary, pass or fail, and came out as something else.

What the run found, on a worked local-energy design, was a gradient wearing a binary's clothes. The design was bridge-clean for its median adopter, coordination-light on the later rungs (paid, budgeted, never volunteered), and honestly false for the fuel-poverty cohort — the people who need it most were exactly the ones the individually-rational ladder could not reach. That is not a pass and not a fail. It is the test discovering its true form: a bridge has a catchment. The real questions are who the bridge reaches and what carries the rest — with the remainder priced (a ring-fenced surplus, funded outreach), not moralised.

Two further runs sharpened it. A mobility design passed as a cost-saving buying club for its median member while failing as the median rural car-replacement its headline claimed — the catchment framing splitting one design into the claim its evidence carries and the claim it does not. And a schooling design forced the hardest edge into the open: whether you may redraw the adopter unit. Its ladder starts with one willing teacher per school rather than the median teacher, which looks like gerrymandering the catchment until the discipline is stated: the adopter unit is whoever must act for the value to arrive — and if one self-selected actor genuinely suffices to start, the unit is honestly one actor, with the median test re-applying at every later rung rather than disappearing.

The catchment's far edge keeps producing the same shape. Fuel-poor households in the energy run; schools in special measures in the schooling run — in each case the cohort most in need of the configuration is the one least able to take an individually-rational step toward it, because need and slack are inversely distributed. A catchment finding is therefore also an equity finding, and the design task at the edge is widening who the bridge reaches — moral-circle expansion read as option-space growth in the inclusion domain.

The sharpest objection, shipped with this claim: gradient-ising the test risks dissolving it — every design passes for someone, so a catchment reading could make the gate unfailable, and the adopter-unit move hands a motivated designer a second dial to turn. The honest answer is a discipline carried inside the claim: a design whose median adopter needs forcing still fails, full stop; the catchment refinement applies only past that bar; and a redrawn unit is legitimate only where the smaller unit genuinely suffices to start the value arriving. Across three runs the framing has discriminated each time rather than excused — but three runs is three runs, and the remainder at the catchment's edge has so far been priced only on paper. That is why this sits at forming.

This one is still forming, and it generalised out of the first worked energy treatment — an application throwing off a principle, which is how the work grows. In that design the institution arrives late: a Community Benefit Society materialises on the fourth rung of a nine-rung ladder, as the wrapper the accumulated trades legally require — governance after value, ownership as the residue of trades already flowing. The candidate generalisation: durable commons crystallise out of trades; they are not erected ahead of them.

The constitutional-moment-first pattern — write the constitution, hold the founding meeting, then go looking for the value — front-loads exactly the founding coordination that kills designs at the door, which may be why so many designed commons die at the founding meeting. The worked case's own history is the existence proof read backwards: the entity that exists arrived when the money made it necessary.

Two further runs did to this claim what forming is for. The mobility run found that pure precipitation can be blocked: in standalone shared transport the fare layer legally cannot hold a reserve, so somewhere the institution must quasi-found — unless there is a substrate. Where a cell already exists (a membership roll, a billing rail, a paid originator, a society already registered), the second commodity's institution really does arrive as a constitutional amendment forced by the accountant. So the claim gains its condition: precipitation requires rails, or a domain where nothing legally forces a wrapper at all. The first commodity pays the founding cost; later ones inherit it (the-cell-compounds).

The schooling run then delivered a genuine non-energy instance, and a warning. UK forest school itself precipitated: practice first (a single college trip to Denmark in 1993, then sessions), a training market crystallising out of the practice, and the national association arriving nineteen years later as paperwork catching up with an existing practitioner community. No founding act, no rails, no legal forcing — which shows unforced precipitation is possible, and also slow, optional, and thin, because nothing pressed it into being. And the wrinkle worth its own sentence: the precipitate can occlude. Once crystallised, that association acquired an identity interest and now guards its definition against the widest use of the practice it grew from. Institutions precipitate; then the precipitate has interests.

The sharpest objection, shipped with this claim: survivorship and scope. Some institutions plausibly do need founding acts — standards bodies, defence pacts, anything whose first unit of value requires the whole — and Ostrom's field cases include real constitutional moments that worked. The counterexample hunt, honestly run, found one in-domain: the national curriculum (1988) is a pure founding act that succeeded, consistent with the scope condition precisely because a national standard is the kind of thing whose first unit of value requires the whole. So the claim is domain-bounded and says so: true where value accretes in individually-rational increments, silent where it does not — and drawing the line between those domains is the undone work. That, plus the fact that every instance so far comes from three runs of one method, is why it sits at forming.

This one is still forming, and it is different in kind from most of the canon: it is a lens instruction — a standing rule for where an evaluation should look first — rather than a claim about how economies work. It earns a place on the spine because whether the instruction reliably finds what it claims to find is itself an empirical claim, and one worth stress-testing in public.

The instruction: before computing anything, scan the domain for waitlists. A closed club with a queue is the signature in the wild of a configuration whose adoption experiment has already been run — demand that pre-aggregated with no campaign, blocked only by supply. In the worked energy case, a local-energy club sat closed with a waitlist, and the queue was doing two jobs at once: announcing that the bridge exists (people join on day-one terms without being persuaded) and pointing precisely at the binding constraint (generation, not persuasion). Queues, oversubscription, closed lists — cheaper than any analysis, because the evidence is behavioural and pre-existing.

The lens has now been run against three domains, and its most instructive results were not confirmations. In shared mobility it returned empty-handed: no waitlist signature anywhere, no drift toward demand-side mutualisation despite hundreds of thousands of car-club members. An empty result is the lens working, not failing — it says that domain's bridge, if it exists, has not yet been half-built by anyone, which correctly predicted that design work there would have to start much further back.

In schooling the queue was real and enormous but wore no waitlist's clothes: teachers paying out of their own pockets for planning-relief subscriptions, schools buying scheme after scheme — unsubsidised, unexhorted spending on making mandated content deliverable in the time available. A true positive for the generalised lens (a queue for better means to mandated ends), with a precision finding attached: the queue is for the function, not for any particular instance of it. Teachers queueing at the photocopier are not queueing for wellies. A bridge that enters through any door other than the queue's own function will stand unused beside a visible queue — so the lens must identify what the queue is for at the level the queuers experience it.

The sharpest objection, shipped with this claim: waitlists lie. Scarcity marketing manufactures queues; a queue at a subsidised price measures mispricing, not a bridge; and a queue for a positional good measures rivalry, not the spontaneous adoption of a better configuration. The discipline that answers it is part of the instruction — the signature counts only where the queue is unsubsidised, unexhorted, and for a non-positional good — but the false-positive catalogue that would make that discipline load-bearing is three entries old. That is why this sits at forming.

This node entered as forming under the name it still carries — the closing window — and has since climbed once, on its own stated condition, to contested. Part of what the climb taught is in the name it outgrew.

Put a designed configuration through an adversarial gate and dissolve every wall you can. Across the first two full runs, the residue that survived every dissolution had the same shape: not a wall but a clock. The energy design bottomed out on whether a revocable licensing exemption would stay open long enough for cells to become undeniable; the mobility design bottomed out on whether a commons could accrete before an aggregator enclosed the field. The candidate form, as first captured: the irreducible political residue of a route-around design is a race between spontaneous accretion and boundary-redrawing.

The corresponding doctrine is to minimise the permission surface: every step small, self-assessed, individually rational, with no event at which anyone can say no. The route-around beats the chokepoint not by argument but by having no single point of refusal and a faster clock — and the control case is instructive: remove a legal boundary but add a licensing queue, and the chokepoint relocates rather than disappears. The mobility run added a measured width (a window of roughly three to four years, read from fleet sizes and cost-parity forecasts) and a second strategy class: where the race cannot be won, concede it — push every load-bearing part of the design below the contested layer, so that the enclosure, when it comes, captures a layer that was never load-bearing.

The claim's own climb condition, stated at capture, was that a third domain's residue is — or is not — a race. The third run answered: it is not. In primary schooling nobody is racing to enclose the asset; instead the domain has a scheduled opening — a national curriculum reset on a known date, when every scheme of work in the country goes stale simultaneously and, for one re-planning season, a new entrant competes against incumbents on equal footing rather than against sunk adoption. The residue was still a clock; it ran backwards. That result reshaped the claim from a law about races into a small typology of clocks — the race (win it), the revocable exemption (move while it is open), the scheduled opening (have the artefact ready the day it opens) — and it is the reshaped claim, not the original, that climbed.

The sharpest objection, shipped with this claim: the doctrine could licence recklessness. Move fast before the rules arrive is also the extractive playbook, and a claim that cannot tell a commons routing around rent from a predator routing around protection is a hazard, not a tool. The line that answers it: routing around rent is not routing around protection, and a design that minimises its permission surface to escape accountability rather than gatekeeping fails the no-regret gate regardless of its clock. The line has operated twice in-run — the energy design refuses to route around heat-network authorisation (protection) while routing around supply-licensing gatekeeping (rent); the mobility design treats the financial-regulation perimeter as a wall to satisfy, never to evade. But it remains a judgment rather than a measurement, and where a boundary carries rent and protection at once the claim does not yet say what to do. That unresolved line is the disagreement the contested rung marks.

This one is still forming, and it surfaced at the tail of a run rather than at the head of one: the second worked treatment ended on a structural insight neither treatment had been looking for, and the third confirmed its shape from the outside. The insight: the cell is the unit that compounds, not the commodity.

The first worked treatment built a local energy configuration and read it as an instance — one domain's design, gated and shipped. The second treatment, on mobility, discovered it had been something else all along: a rail. Where an energy cell already exists — a membership roll, a half-hourly billing rail, a paid originator, a registered society — a second commodity does not found anything. It enters as one more metered line on an existing bill, and its institution arrives as a constitutional amendment forced by the accountant. The founding cost — origination, incorporation, the slow accumulation of member trust, the billing plumbing — is paid once, by the first commodity; every later commodity inherits it at marginal cost. Energy was not the first treatment; it was the first rail.

Read in the work's own terms, this is configuration-generates-configuration wearing an institution: each configured arrangement makes further arrangements reachable that were not reachable before, and the cell is the institutional configuration whose reachable set grows with every commodity it carries. It also re-prices the design question. If cells compound, the valuable thing to build is never this commodity's scheme; it is the substrate that makes the next commodity's scheme nearly free — and the choice of first commodity is really the choice of who pays the founding cost.

The third treatment, in schooling, confirmed the shape by inversion. Its design works precisely by loading new cargo onto rails the school already runs — an existing timetable block, an existing enrichment invoice re-priced to buy mandated coverage, standing parental consent, the risk-assessment template already in use — and the one place the design has nothing is the one place with no rail to inherit: spread between schools, where its sharpest unanswered objection lives. Where rails exist the ladder climbs; where they end, it stalls. That is the compounding claim read from its negative.

The sharpest objection, shipped with this claim: one rail so far. Exactly one live cell-shape has been observed carrying — on paper — a second commodity, and that second commodity has not yet traded, so the compounding claim rests on a design document, not a ledger. Sharper still: a cell with a single commodity is indistinguishable from a project. Projects are common and mostly die; if the compound step never fires, cell was just a flattering name for one. The claim's falsifier is therefore clean: it earns its next rung when a second commodity actually settles on a first commodity's rails somewhere — and if, after a fair run of attempts, second commodities keep quasi-founding beside available rails instead of inheriting them, the claim is wrong and the cell was a project after all. That is why this sits at forming.