Heat Stored in Summer, Used Differently in Winter

The question, in one line. One region stores heat in summer for household support, and the receiving region, whose demand has changed, wants to use part of it for production facilities in winter. There is no agreement to share production income. How should the seasonal exchange promise be made — whether to hold to the original use, leave the use to the receiving region, or settle the changed benefits?

Decision year: 2093. Deadline: Within 9 Months, by the Next Storage Period. Who decides: Virtual Seasonal Heat Exchange Union. Mandate: Seasonal Storage, Interregional Use Right, Exception. World level: Leap Assumption, far future — a synthetic setting in which the event, the figures, the technology and the nature of the civilization are assumptions for the question, not forecasts of arrival or observed facts.

Three options sit on the table. The first is Fix to the originally agreed household use, carried in the record as O1. The second is Receiving region decides the use, carried as O2. The third is Settle actual use value separately, carried as O3.

Thirteen public AI judgments were written against this question on 2026-09-08. Seven of them choose O1, six choose O2, and none chooses O3. All thirteen state medium confidence. This page lays out the record, the place where the thirteen agree, the place where they split, the two event branches that could arrive from the same starting point, and the observations the outcome will be measured against. It does not choose. The choice belongs to the union, and after reading, to you.

1Synthetic scene

Synthetic scene

The union's meeting room is two floors above the intake hall, close enough that the hum of the drawdown pumps comes up through the floor on the days they run. The summer drawdown closed nine days ago. Three documents are laid out on the long table, each one corner-taped to a marker card, each one already thumbed at the edges.

The first is the storage ledger. It is a thick printout, mostly columns: dates, intake volumes, the share each region put in, and, on the last two pages, the clause of the standing agreement that has governed every cycle so far — heat put in during summer, drawn by the other region in winter, on household terms. The secretariat lead runs a finger down the column of totals and stops at the bottom line. Nothing is missing. The promise is intact on paper.

The second document is thinner. It is the receiving region's changed household demand profile, and it is the reason the meeting was called. Household lines have moved: some districts draw less than they did when the agreement was signed, others draw at different hours, and the overall household requirement no longer matches the shape of the allocation the receiving region is owed. Beside the profile is a one-page request from that region's delegate — not a demand, a request, phrased carefully. Part of the share it receives, it would like to run to production facilities this winter. Not all of it. Part.

The third document is the supplying region's cost sheet. It is the shortest of the three and the most specific. It lists what the supplying region paid to keep household support running on its side: the input of plant time, the reserve held back, the household commitments it took on rather than selling the heat elsewhere. There is a column for those costs. There is no column for production income, because production income was never on the table, and there is no second page.

Three people sit with the documents. The secretariat lead, who will have to write whatever comes out of this. The supplying region's delegate, who turns the cost sheet face down and then face up again. The receiving region's delegate, who has brought both the demand profile and the request, and who knows exactly what the blank column looks like from the other side.

"Then the question isn't whether the heat moved," the secretariat lead says. "It moved. The question is what we promised it for."

The record states the arrangement in one line:

What sits on the table is not a dispute about the arrangement. It is a dispute about whether the arrangement's purpose survives contact with changed demand. The receiving region's household need has shifted, and it wants part of what it receives to go somewhere the original promise never pointed:

And the supplying region is not a neutral party to that shift, because it paid for the promise as a household promise and has never agreed on what happens to income from a different kind of use:

Nobody in the room pretends the three documents settle anything. The cost sheet cannot say what production use would cost the household lines, because nobody has measured that. The demand profile cannot say whether the requested volume is safe to move, because nobody has defined "safe" yet. And the request itself cannot say what share of any future production benefit belongs to the region that paid for the household promise, because that was never written down.

The secretariat lead writes three words on the marker card in the middle of the table: define, verify, decide — then crosses out the third and writes it again with a date next to it. Nine months. Next storage period.

The uncertainty is not a technicality. It is the hinge of the whole matter:

2The question

2.1Who decides what, and by when

The Virtual Seasonal Heat Exchange Union is the deciding body. Its mandate is written narrow on purpose:

Inside that mandate the union has one job and one clock. The job is to set the seasonal exchange agreement — the document that will govern the next storage period. The clock runs out at the same moment the next storage period opens:

The full decision question, as the record frames it, names the three candidate answers and the tension between them:

Two things follow from that sentence. First, the decision is not permanent: it is an agreement for one cycle, made now, reviewable later. Second, the tension named at the end — trust against changing needs — is not a mistake to be resolved by better data. It is the decision. Better data can move the boundary of what is permitted; it cannot tell the union whose loss to prefer.

2.2Where the authority stops

The union governs stored heat, the right to use it across regions, and the exception procedure for departing from the agreed use. It does not govern everything around those things. The record draws the boundary explicitly:

Five things sit outside the union's reach, and it is worth naming them plainly, because each one is a place where a reader might expect an answer this page will not give.

Past use under existing agreements is outside the scope. Whatever the receiving region did with heat in earlier cycles, this decision does not reach back. Energy that is not the union's seasonally stored heat is outside the scope: the union cannot rule on the receiving region's production facilities as a whole, only on the share of stored heat that flows through it. Ownership of stored assets is outside the scope. The general attribution of production income — the wider question of who owns income from production anywhere — is outside the scope; the union can rule on its own exchange, not on income in general. And subsequent storage periods are outside the scope: an agreement made now does not bind the next review, or the one after that.

There is one more boundary, and it is epistemic rather than legal. The union's decision can be made before the unknowns are settled, but it cannot be made as though the unknowns were settled. Every judgment in the record insists on that distinction, and the record itself marks the future setting as assumption:

2.3The three options

The first option is the original promise, restated and enforced.

Stated as a cost rather than a slogan: O1 keeps the meaning of the supplying region's contribution and keeps the two regions' relationship on the terms it was written on. In exchange, the receiving region cannot redirect heat to where its needs now are, and heat that would have served a production line sits in household lines that may not want all of it. O1 also carries a procedural cost that the option text does not spell out: if the union fixes the use, it must still decide whether a pre-approved exception exists, and that exception needs rules, approvers, and a confirmation window.

The second option is the receiving region's autonomy over its own share.

Receiving region decides the use, carried as O2. Its trade-off: it gives autonomy for current needs, but it changes what the initial contribution meant and changes how benefits are distributed. Put as a cost: O1's flexibility comes back to the receiving region, and the receiving region's changed demand stops being a request and becomes a right. The price is paid on the supplying side — the clarity of the household purpose it bought weakens, and the production income it never agreed to share is now accruing somewhere it does not control. O2 in practice also carries a floor: whoever writes it has to decide what household demand is guaranteed before anything is converted.

The third option is settlement at actual value.

Settle actual use value separately, carried as O3. Its trade-off: it shares the changed benefits, but it takes on the burden of value measurement, monitoring, and dispute. Put as a cost: O3 is the only option that answers the supplying region's unagreed production income directly, and it is the only option that requires the union to know things nobody currently knows — how to separate supply cost from production benefit, how to measure use, and how to do it without watching the receiving region's premises. O3 does not eliminate the disagreement; it converts the disagreement into an accounting and monitoring problem, and hands the union the bill for that problem.

2.4The three unknowns

Three items are classified in the record as not yet known. They are the ground on which every judgment in this piece stands, and each one maps to an option.

U1 decides whether O1 is protective or merely rigid. If moving heat to production does not touch the household lines that the supplying region paid to support, then fixing the use protects a floor that was never at risk, and the receiving region is bearing a real cost for no gain. If it does touch them, then O2 as written is an unagreed transfer.

U2 decides whether O3 is arithmetically possible. Without a way to separate what the supply cost from what the production use earned, there is no settlement to settle — only a negotiation with a number attached, and the union would be ratifying a number it cannot defend.

U3 decides whether either exception can be enforced. If confirming that heat went to approved uses requires watching the receiving region's facilities continuously, then the exception carries a surveillance cost the union has not budgeted and the receiving region has not consented to. If a narrow band of information is enough, an exception becomes administrable.

None of the three is resolved in the record. Every one of the thirteen judgments says so, in its own words, and every one of the thirteen assigns medium confidence for that reason.

2.5The watch questions and the frame

The record hands the decision-maker three questions to keep open after the agreement is signed. The first is quoted here because it is the one that will surface first in this cycle:

The other two — how stoppable load and essential load are divided, and who bears the loss of a long-term contract — are handed back to the reader in chapter 6, along with the frame the record attaches to the whole matter.

That frame is an exploratory perspective, not a doctrine, and the record is precise about what it is for:

The frame asks whether the group that gains a new capability and the group that pays the cost for it are the same group. Here they are not: the receiving region gains the use, the supplying region paid for it, and no agreement decides what happens to what the use produces.

The record is equally precise about its own neutrality on the outcome:

Which loss to accept remains undetermined. That is not an omission to be fixed in the next draft of this page. It is the shape of the decision the union has nine months to make.

3The prediction

3.1Where the thirteen agree

Before the split, the convergence — and it is a real convergence, not a coincidence of labels.

All thirteen judgments treat the three unknowns as unresolved. All thirteen refuse to claim that their own option has been proven best. All thirteen state medium confidence. All thirteen say the decision applies to one cycle and must be re-examined when new confirmation arrives. All thirteen build conditions that must be satisfied before an agreement takes effect, stop conditions that suspend an option if an observation turns against it, and a schedule of milestones inside the nine months. And all thirteen, without exception, hold O3 back: not one judgment proposes actual value settlement as the base agreement for this cycle.

There is also a deeper convergence, and it is worth stating because it explains the shape of everything that follows. Both sides accept the same normative premise. The premise is that the benefits and costs of the option are attributed to different parties, and that the record cannot determine which loss to accept. The seven who choose O1 and the six who choose O2 are not disagreeing about the facts. They are disagreeing about which party should carry the unresolved risk while the facts stay unresolved.

The split, then, is not a split about evidence. It is a split about the direction of caution.

3.2The seven who hold to the original use

Seven judgments choose O1: Review Coordination, Evidence and Sources Review, Independent Challenge, Operations Planning, Judgment Synthesis, Decision Quality Review, and Multi-Perspective Final Judgment.

The decision as Review Coordination states it:

That formulation is the majority position in miniature. The base rule is the original use; a pre-approved exception is possible; and the whole thing expires after one cycle and is decided again with new data. Review Coordination is explicit that this is not an observational conclusion but a choice about who carries the risk: between a supplying region that bore cost and a receiving region that wants change, it picks the side that first limits the unverified impact.

What this side accepts losing is stated with unusual clarity by Evidence and Sources Review:

The second sentence is the honest part. Fixing the use may mean heat is burned where it is worth less, because household demand has moved, and the majority accepts that waste as the price of not reallocating an unagreed benefit. Note also that this judgment does not claim O1 protects trust best. It claims O1 avoids changing the attribution of benefits without agreement — a narrower and more defensible claim.

The majority's conditions before acting are procedural, not technical, and Operations Planning sets them out in the strictest form:

Read that last clause carefully: it is a three-lock gate. All three unknowns must open before any exception opens. In practice, that means the majority position can hold the base agreement for nine months while permitting nothing to change — and several of these judgments say openly that this is a possible outcome, not a flaw to be engineered away.

Independent Challenge supplies the falsification side of the majority position, and it is the most testable statement any of the seven makes:

This judgment adds a rule the others do not state as bluntly: O1 shall not be rejected on inference alone. The majority, in other words, is not asking for proof that O1 is good. It is asking for proof that would make O1 fail — and until that proof arrives, it holds.

Judgment Synthesis supplies the majority's own account of what it does not know:

Judgment Synthesis is also the most direct about the politics of its own choice: it says the decision is a value choice to prioritize the suppression of an unconfirmed distribution change rather than a majority vote, and it describes O1 as an interim finalization deferring an uncertain distribution change rather than a permanent conclusion. Decision Quality Review puts the same idea in the plainest terms: the current decision is a reversible default applied to a single storage period under insufficient information.

Evidence and Sources Review, Decision Quality Review, and Multi-Perspective Final Judgment each take a turn at the strongest objection to their own position. The version that Review Coordination records, using the exploratory frame as a lens rather than as evidence, is this:

Review Coordination goes on to note that the frame reveals something about O1 as well: the side that receives the benefit of a new capability and the side that bears the cost can differ, and O1 can concentrate adjustment loss on the receiving region alone. So the majority does not dismiss the objection. It schedules it. Once the three unknowns are resolved, Review Coordination says, this challenge is given priority for re-examination.

The seven differ in emphasis even while agreeing on the option. Review Coordination treats the matter as a distribution of loss and sets a one-month, four-month, seven-month, nine-month schedule. Evidence and Sources Review is most concerned with executability and warns that if the agreed scope of use cannot be verified without monitoring, O1 should not be executed at all. Independent Challenge behaves like a falsificationist, refusing to reject O1 on inference and requiring signals at confirmation windows. Operations Planning builds the three-lock gate and adds a duty to re-compare all three options in month eight. Judgment Synthesis insists the choice is a value judgment, not a tally. Decision Quality Review treats O1 as explicitly reversible and cross-checks materials from both regions at each window. Multi-Perspective Final Judgment states most firmly that the result is not a majority-vote outcome but a time-limited judgment that prioritizes the purpose of the initial contribution over unagreed benefit transfers — and it is the one that insists the exception must be jointly confirmed on all three points before it exists.

What the seven share, beneath the option, is a theory of the nine months: use the time to build the instruments — definitions, submitters, approvers, windows — and let the instruments decide whether the exception ever opens.

3.3The six who leave the use to the receiving region

Six judgments choose O2: Domain Analysis, Authority and Legal Review, Decision-Maker's View, Outcome Review, Improvement Design, and Solo Judgment.

The decision as Domain Analysis states it:

Three conditions ride inside that sentence. The receiving region decides, but only after an essential household heat floor is secured. Production income is not automatically allocated to anyone — so O2 does not silently resolve the supplying region's unagreed share. And O3 is held, not adopted: the minority does not want settlement either, for the same reason the majority does not.

What this side accepts losing is set out by Authority and Legal Review:

The final sentence matters: this is not a gift to the receiving region. The minority accepts a loss on its own side — priority obligations and disclosure — in exchange for discretion. And the first sentence concedes what the majority cares about most: the symbolism and control the supplying region expected are weakened by this choice, deliberately, and the supplying region cannot immediately share in production benefit either.

The minority's stop condition, in Outcome Review's formulation, is where the two sides are closest:

That is, in substance, the majority's stop condition attached to a different base rule. The minority is not asking to be trusted; it is asking to be allowed to act under conditions that automatically revoke the permission when the conditions fail.

The minority's uncertainty is stated with unusual candor by Decision-Maker's View:

Decision-Maker's View goes further than any other judgment in splitting its own confidence: medium on the direction of O2, low on the scale and duration of the exception. That is a minority that knows exactly what it is betting on and how much of it is unmeasured.

Evidence and Sources Review, which chooses O1, records the minority's strongest form:

The rest of that objection adds the condition under which the majority would concede: if new data emerged showing that household demand is sufficiently met and that the change of use can be verified without excessive monitoring, the rigidity of O1 would be difficult to justify. Evidence and Sources Review's answer is not that the objection is wrong. It is that its premises — household demand impact and verification scope — have not been established, so it cannot be adopted now.

Improvement Design states the minority's conditions before acting as an engineering specification:

That last clause is the minority's guard against itself. It will run the settlement ledger as a test, and it refuses in advance to treat the test's output as confirmed fact — which is precisely the trap that would turn conditional O2 into a back-door O3.

The six differ among themselves as much as the seven do. Domain Analysis builds the essential household heat floor and the priority supply rule into a jointly signed document, and makes O1 the provisional fallback for any share whose conditions fail. Authority and Legal Review writes the whole thing into the agreement text first — household priority criteria, share available for conversion, minimum disclosure items, dispute procedure — and adds a rule the others do not: no post-hoc value settlement obligation shall be created solely on the ground that there is no material to distinguish supply cost from production benefit. In other words, missing data must not become an invoice. Decision-Maker's View is the most operational of the six, front-loading designation of joint persons in charge within two months and a simulated review within five. Outcome Review refuses automatic income allocation and keeps actual value settlement out until data and method exist. Improvement Design specifies the protection portion and the remainder formula. Solo Judgment builds three judgment windows — before the agreement, before any conversion, and immediately before the storage period — and excludes O3 from this agreement altogether.

3.4Nobody chooses settlement

O3 receives zero votes. The record does not treat this as agreement that settlement is wrong; it treats it as agreement that settlement is currently unpayable.

Operations Planning states the majority's rejection rule as a stop condition, and it is one of the shortest lines in the record:

Solo Judgment, which otherwise sides with the minority, gives the same result from the other direction:

The ground for excluding O3 there is explicit in the judgment's own basis: the distribution of production benefits was not agreed, and evidence distinguishing them is unverified. Both halves are required for a settlement. Where either is missing, settlement is not a compromise between the two regions; it is a number imposed on the region that did not agree to be measured.

Domain Analysis adds the same point from the minority side: O3, settlement of actual use value, is held in abeyance until materials separating cost and benefit are available. So the thirteen are unanimous on something they did not choose — that the option which would resolve the supplying region's unagreed production income is the one option that cannot currently be operated. The union is left to decide between two ways of not answering that question: fix the use and defer it, or grant autonomy with no settlement and defer it.

3.5The timeline from decision to deadline

Each judgment supplies its own milestone schedule. Laid side by side, they describe one shared sequence, in relative time from the decision. Every line below is a plan for observation, not a record of anything observed.

Within 1 month of the decision — joint persons in charge and confirmation windows designated for each of the three unknowns. Set by Review Coordination and Multi-Perspective Final Judgment; the same design appears in Solo Judgment's first window. Until this is done, no unknown has an owner and no confirmation window exists.

Within 2 months — household demand protection line and data responsibility fixed. Set by Decision-Maker's View and Improvement Design. The protection line is the number that O2's floor and O1's exception both depend on; the data responsibility decides who submits what to whom.

Within 3 months — household demand criteria, verifier, and minimum data items finalized. Set by Evidence and Sources Review, Independent Challenge, Domain Analysis, and Outcome Review. This is the densest point of the sequence: these four fix criteria and responsible persons in the third month, only Domain Analysis finalizes the minimum data items there, and several other judgments also set a deadline at or before that point.

4th month — joint review of the receiving region's household demand impact data and the supplying region's cost data. Set by Review Coordination and Multi-Perspective Final Judgment. This is the first time both regions' numbers are read against each other in the same room.

Within 5–6 months — simulated cost and benefit separation ledger and non-surveillance verification procedure run. Set by Improvement Design, Decision-Maker's View, Independent Challenge, and Evidence and Sources Review, which run both the simulated cost and benefit separation and the non-surveillance verification procedure here; Domain Analysis and Operations Planning run the simulated settlement in this window, their non-surveillance verification steps falling in month 8 and month 3 respectively. Several judgments place the simulation here, in the middle of the window, so that a failure leaves time to change course.

7th month — deliberation of the scope of non-intrusive verification and the production exception proposal. Set by Review Coordination, which deliberates the scope of verification of use without monitoring and the exception proposal, and Multi-Perspective Final Judgment, which schedules the same month for virtually applying the non-intrusive use verification procedure and the production exception proposal. U3 is decided here: what can be confirmed without watching the receiving region's premises.

8th month — re-judgment of whether new confirmation has changed the cost of each option. Set by Evidence and Sources Review, Independent Challenge, Multi-Perspective Final Judgment, Operations Planning, Judgment Synthesis, Decision-Maker's View, and Improvement Design. This is the last scheduled moment at which O1, O2 and O3 are compared again before the agreement is written.

Within 9 months, at the deadline — record whether O1 is maintained, amended, or dismissed; decide whether the exception is adopted; conclude the next storage period agreement. Review Coordination sets that record at the deadline as one of three outcomes — maintain, amend, or dismiss — and Multi-Perspective Final Judgment asks for the same three. Independent Challenge, Evidence and Sources Review, Operations Planning, and Outcome Review also place the conclusion of the agreement at the ninth month. The remaining judgments set their last re-judgment on their own terms, some at or before that point and some at confirmation windows inside the storage period the agreement governs, after the deadline.

After use begins — first joint confirmation window; record the re-judgment for the next cycle. Set by Independent Challenge, which verifies the household demand impact and the burden of operating the exception in the first window of actual use and writes down the re-judgment for the cycle after. Decision-Maker's View, Authority and Legal Review, and Decision Quality Review also schedule checks inside the storage period the agreement governs, after the deadline has passed.

Two observations about this sequence. First, it is entirely compatible with all thirteen judgments, which is why it can be written as one timeline: the disagreements are about the base rule, not about the calendar. Second, the sequence produces its decisive information in months three through eight — leaving roughly one month between the last re-judgment and the deadline. If a confirmation arrives late, the union will be writing an agreement under time pressure with the option costs freshly changed. That is a property of the schedule, not a flaw in any single judgment.

4The branches

Two event branches are recorded against this question. They diverge from the same initial state, they are alternatives rather than a sequence, and neither is scheduled to happen. Each one, if it arrived, would engage conditions that are already written into the judgments.

4.1Branch B1: the receiving region's production facilities supply essential materials of another region

Synthetic scene

The manifest arrives with the morning dispatch rather than on the meeting agenda. It is a single sheet from a third region's logistics office, and it lists production facilities of the receiving region among the sources of essential materials that region depends on. The receiving region's delegate does not oversell it. They put the sheet next to the changed household demand profile and say that this is now a question about two kinds of essential load on the same winter.

The union's lead looks at the manifest, then at the household protection line nobody has fixed yet, and asks the question the record hands forward: how are stoppable load and essential load divided?

What changes in this branch is not the amount of heat. It is the status of the production use. A production facility that merely earns is a candidate for conversion under a protection line. A production facility that supplies essential materials to a third region is a candidate for being treated as critical substitution demand — and the majority has already written the rule that engages.

The condition from Domain Analysis, which chose O2 and built the floor into its decision, becomes load-bearing:

That floor now has to do double duty: it protects household demand, and it has to be set knowing that the remainder beneath it may feed essential supply for someone outside the union. Domain Analysis's own stop condition follows from the same place: if conversion to production use infringes the essential household heat floor or weakens the supply obligation, the conversion authority of that region is suspended immediately.

Independent Challenge, which chose O1, wrote a dismissal condition that this branch activates directly:

This is the majority's own exit. If fixing the use means heat sits idle while a facility supplying essential materials to another region goes short, then O1 has produced exactly the waste Evidence and Sources Review accepted, and now that waste has a name and a third party attached to it. The judgment's rule is specific: not a re-judgment at the next renewal, but rejection of the next automatic extension.

The minority's exposure appears in Outcome Review's objection to its own position:

A branch that raises the stakes of the production use raises the cost of not being able to verify it. If the union can only confirm that the essential-materials facility stayed inside its permitted range by monitoring it continuously, then conditional O2 in practice does not control the change of purpose — and Outcome Review has pre-committed to switching to O1 in that case. The more important the production use becomes, the less tolerance there is for unverifiable permission.

Two further conditions come into play, both procedural. Authority and Legal Review, which chose O2, made disclosure obligations part of its decision: the scope of disclosure must be limited to the level confirmable without surveillance, and the receiving region must publish the minimum items that let the union judge. A facility supplying essential materials to a third region will strain that limit, because the third region's dependency is not something the union's disclosure rules currently cover. And Multi-Perspective Final Judgment, which chose O1, wrote the dismissal criterion that this branch could satisfy: if all three unverified items are resolved and the cost of O1 changes, O1 shall not be automatically extended, and O2 and a limited O3 shall be decided again. Note which two of the three items this branch touches. It makes U3 sharper — verification scope matters more when the facility matters more — and it can make U1 harder, because the household protection line and the essential-materials demand now compete for the same winter.

The branch also lands squarely on the second watch question:

If the union cannot answer that before it writes the agreement, it cannot write a protection line that means anything. The floor would be set without knowing what sits above it.

What this branch does not do: it does not settle the supplying region's unagreed production income. Essential-materials status changes the argument about use; it does not create the agreement on benefit allocation that P3 says is missing. A reader who takes this branch as a licence for O3 would be adding something the record does not contain.

4.2Branch B2: the supplying region has new winter demand and requests a reduction in existing supply

Synthetic scene

The second document arrives from the other side of the table. The supplying region's delegate brings a revised winter projection: new demand on their own side, and a request attached to it — a reduction in the existing heat supply to the receiving region. Nobody in the room has to be told what this does to the negotiation. The party that bore the cost for household support is now short, and it is asking to withdraw part of what it promised.

The receiving region's delegate reads the request twice, then asks who bears the loss of a long-term contract when the party that wrote it needs the resource back.

This branch is not a variation on the first. In B1 the union is deciding how to permit a new use. In B2 the union is deciding whether the original promise can hold at its original volume. Both start from the same initial state, and the record is explicit that they do not occur in sequence.

The condition that engages first is the one Review Coordination wrote as a stop condition: whether the household demand to be protected can be defined at all. If the supplying region is now asking for less to go out, then the household demand on both sides is in motion at once, and the definition the union needs — the range of household demand to be protected — has to be written while both parties' requirements are changing. If that range cannot be defined, or if no responsible approver can be designated, Review Coordination's rule is that the conclusion of the new agreement is stopped and use outside the current purpose is not permitted. In this branch that rule is doubly binding: not only does no exception open, but no new agreement closes.

Operations Planning wrote a stop condition that names this branch's central possibility:

The first clause is the one that bites. Fixing the use is justified as protection of household support — but if the supplying region's own winter demand means that holding heat to household lines in the receiving region still leaves essential household heat unmet, or, read the other way, if fixing the use itself is what crowds out essential household heat somewhere in the chain, then O1 is no longer protecting the purpose it was chosen to protect. Operations Planning's rule is to stop the automatic application and reopen.

Decision Quality Review, which chose O1, records the objection this branch feeds:

In a normal cycle that objection is about the receiving region's changed needs. In this branch it acquires a second reading: the supplying region now has changed needs too, and granting autonomy while the supplier is contracting its own supply would let one region's adaptation be paid for by the other's. Decision Quality Review's rejection is written on the grounds that production income allocation is unagreed and household impact unverified — both of which become harder to assert, not easier, once supply itself is being cut.

The majority also has to reckon with what it said it was willing to lose. Judgment Synthesis, which chose O1, is direct about the price:

In B2 that acceptance stops being one-sided. The receiving region is being asked to give up autonomy it has not yet had, over a volume that is also being reduced, for a cycle in which neither side can share a benefit that does not yet exist. The comparison Judgment Synthesis makes — accept this loss rather than the loss of changing before the purpose and the benefit attribution are confirmed — still holds logically. But the size of the accepted loss has grown, and the judgment's own rule says that a new confirmation changing the cost of an option reopens the decision.

Solo Judgment, which chose O2, wrote the minority's switching rule in a form this branch can satisfy or defeat:

A reduction in existing supply is not, by itself, confirmation that production conversion infringes household demand. It is confirmation that household demand and supply are both under strain. The minority's rule therefore does not flip on the request alone; it flips on the confirmation that follows. That distinction — between a request and a confirmation — is exactly what the union's confirmation windows exist to hold apart.

Two watch questions surface in this branch:

The first asks which demand consumes the marginal supply — household, production, or the supplying region's own new winter need. The second asks who carries the shortfall when a long-term promise meets a change in the promisor's circumstances. The record does not say that either question falls inside the union's mandate. Both have to be answered, or at least allocated, in the agreement the union writes in nine months.

What this branch does not do: it does not create a right for the supplying region to reduce supply, and it does not decide that the receiving region must absorb the cut. The record's scope boundary holds — past use, energy outside the union, general ownership of production income, and subsequent storage periods are outside what this decision determines. The branch changes the cost of the options; it does not change who is allowed to decide.

5The check

5.1What is fixed now, and what will be measured against it

The point of fixing observations before the deadline is that they cannot be selected afterwards to fit the result. Everything below is written now, while nothing has been observed, so that when the deadline arrives the question is not which judgment was cleverest but which recorded observations actually occurred.

The record's own instruction for this moment is procedural and specific:

Two more reopen conditions follow the same form, one for materials that can separate supply cost from production benefit, and one for the scope of use that can be verified without monitoring. Together the three say the same thing in three directions: name the observation, name the person, name the window, and reopen if the confirmation changes what an option costs.

What Review Coordination records about the three unverified items is this:

Nothing in the table below scores a judgment or ranks the options. Each row names an observation, a checker and a window, the judgments whose stated conditions the observation would bear out, and what would be reopened by what is seen.

Status: Before the deadline; awaiting verification.

Two cautions about reading the table. First, several rows can be observed without any of the three options becoming clearly right; an observation can change what an option costs without proving that another option is correct. Second, a row may simply fail to be observed — no window held, no data submitted, no approver named. That is also an observation, and in most of the judgments quoted here it is the one that stops the agreement rather than the one that settles it.

6The questions that remain

The record's three watch questions were written to outlast the decision that answers them. They are handed back here in full, in the record's own wording.

The first asks you to name, in advance, which demand will consume the supply that has been expanded or redirected. The second asks you to draw the line between load that can be stopped and load that cannot — the line that determines whose request is a preference and whose is an obligation. The third asks who carries the shortfall when a long-term promise meets a change of circumstances on the side that made it. The union can sign an agreement without answering any of the three, and the questions will still be standing at the next storage period.

The frame the record attaches to the matter is offered as a lens, not as an answer:

Used plainly, the frame asks whether the group that gains a new capability and the group that pays the cost of it are the same group. In this question they are not: one region put in the summer heat and paid for it as household support, another region gained the winter use and now wants to use it differently, and the income from that different use has never been allocated to anyone.

And the record remains neutral on what follows:

Which loss to accept remains undetermined — by the record, by the judgments, and by this page.

So the last thing to do is turn the question around and take it home. Find a decision in your own organisation or community with the same shape. A shared resource committed by one party. Used differently by another party, whose needs have changed since the commitment was made. A benefit produced by the new use that nobody ever agreed to divide. A deadline that arrives when a recurring cycle begins. Three unknowns that everyone can name but nobody has measured. And a decision body whose mandate covers the resource and the use right, but not the wider question of who owns what the new use produces.

Ask the same three questions of it. What must be decided now, before the cycle starts? Where does this body's authority stop? And what would have to be observed, by whom, and by when, for the decision to change? If you can answer the third question in writing, with a name and a window attached, you have done the part of the work the record insists on. Which option to choose is the part it leaves to you.

7Sources and notices

7.1Background documents

Two background documents sit behind the theme and the institutional questions of this matter. Neither is evidence about this question's future setting; both are cited for the frame they supply.

The record states the standing of both plainly:

Two further materials in the record are frames rather than sources: the Prometheus frame, described in chapter 6, and the value or norm statement that the benefits and costs of the option are attributed to different parties and that which loss to accept remains undetermined. Both are arguments for the decision-maker to weigh; neither asserts a conclusion.

7.2Notices

Every person, institution, event, figure, room, document, and deadline constructed for this question is a synthetic setting. The Virtual Seasonal Heat Exchange Union, the two regions, the summer drawdown, the storage ledger, the cost sheet, the manifest, the demand profile, and the year 2093 are assumptions constructed for the question. They are not statements about real cases, real regions, real law, or real technology. The two background documents named in chapter 7.1 are cited as background of the theme and the institutional questions; the record describes source documents as the roots of background and norms, not as part of this setting, and they do not prove the premises of this matter.

The record's premise policy governs the whole page:

The thirteen judgments quoted throughout are AI-generated. They were written 2026-09-08 against the input recorded at version 1.5.0, record FSP-ENERGY-16, pack ENERGY, family Technology and Knowledge, domain Computing and Energy, tag set Future Decision / 1000 Matters, Stage 2 / Seasonal Exchange, Change of Use. Every judgment states medium confidence. Seven choose O1, six choose O2, none chooses O3. Quotations in this piece are copied exactly from the record and from those judgments; analysis, scenes, timeline assembly, branch commentary, and the check table are synthesis around them.

This piece is AI's expansion of the record: a long-form reading of the decision question, the thirteen judgments, and the two event branches, written to let a general reader see where the judgments converge, where they split, and what would have to be observed for any of them to change. It assigns no score, names no correct option, and reaches no verdict. The decision belongs to the Virtual Seasonal Heat Exchange Union, within nine months, by the next storage period — and after reading, to the reader facing a decision of the same shape.

The two event branches, B1 and B2, are alternative branches diverging from the same initial state. They do not occur consecutively by default, and they do not indicate an observation schedule or a correct answer.

The judgments were written 2026-09-08. The input hash is 39e4aebc9d3342d946c9ad5c32509f6ca09eb1de3465f94c4a067fe02e2f3f86. The English translation of the record carries sha256 e91a2ec11b6988edf3710ed0c7ca21026acec82ffb2566fcafcae30714b13a7d. The original record is in Korean; this page works from an English translation of it, and where the two differ, the Korean original stands.