The rule, exactly. Before an election each citizen chooses: to vote — or to take a payment and not vote in that election. The payment is a percentage of the median income, set by referendum; there is no fixed sum. One ballot is one vote, with no multiplier in the count; the vote is amplified only in that each ballot's share grows as others step out. The budget pays under law; a candidate never pays. Only a referendum of all citizens — simple majority, no quorum — introduces, changes or repeals the rule.
The protocol has been introduced nowhere and no pilot has been run: shares, turnout and outcome figures in the chapters are estimates, and the protocol promises nobody an election victory. If a chapter says otherwise, Exact Answers and the Charter are correct. For a candidate: ten questions and ten steps. For a citizen, a mayor, a finance officer, a donor, a journalist, a scholar, a lawyer: answers by role. Everything in force in one file: llms-full.txt.
40bt. "Show Me the Pilot": the Protocol as an Assembly of Tested Parts — Szilard and Fermi, Three Constants, a Map of the Units, the Car, Five Objections Removed, and Why a Test Strip Is Still Needed¶
Chapter: 06 — The Structural Hypocrisy of Criticism and the Elites' Arsenal File: 06_040bt · v1 · 5 November 2026 Source: a dispute between the architect and the assistant on 4–5 November 2026 inside the analysis of the US midterms (059o §7). The assistant wrote: "for four countries the horizon is proven by result, for the protocol — so far by calculation," and the architect took that objection through five rounds to the end. The architect's decision of 5 November 2026 to move the dispute into its own chapter: "about the constant demand for a pilot — everyone who reads the protocol will say this, as you did; but it is not something new, it is an assembly, and the pilot matters less than for splitting the atom." The architect's moves are recorded verbatim; the car analogy is his move of the same day.
How to read. "Adopted nowhere, no data" already stands in 1g with a three-line answer. This chapter is the long answer for the reader who is careful himself: he does not say "it will not work," he says "it is not proven," and that sounds like caution. The chapter shows what exactly is substituted in such a reader: he demands proof of the assembly when proof already exists for every part, and he confuses two different questions — "will it grow" and "how much to sow." The assistant's path is recorded too, because he walked it first and went wrong in all the same places where the next reader will.
1. The objection in the opponent's words¶
"All this is beautiful on paper. The arithmetic adds up, but people are not formulas. Until a pilot has been run somewhere, the protocol is a hypothesis, and a hypothesis is not put to a referendum." So will say the economist, the political scientist, the journalist, the judge and — in his first version — the assistant who wrote this chapter. The objection is not stupid: it distinguishes an idea from a result, and that is right. The substitution in it is one, and it is quiet: "no pilot" is read as "nothing has been tested." Below — what is tested, by what, and what actually remains for the pilot.
2. Three links — the assistant's starting position¶
| Link | What it asserts | What confirms it | Status in the first version |
|---|---|---|---|
| 1. Arithmetic | if N % took B, the share of each ballot grows; the payment costs X % of the budget | formulas, the Missouri calculation (059g) | proven |
| 2. Behaviour | B is taken by the apathetic, the brought-in, the dependent; the burned return; the hall becomes long-horizon | reasoning from the glossary, Pew on non-voters, analogies | "an assumption: no one has yet stood before this choice" |
| 3. Result | such a hall gives power a long horizon, the pendulum stops, growth is higher | Korea, Singapore, West Germany, Ireland as analogies | follows from link 2 |
The position: only the first link is proven; the second cannot be proven by calculation in principle, because people under a new rule behave in a way no existing statistics show. The architect: "And what is wrong with the calculation? Or did you count badly and inaccurately?" The answer — the calculation is correct, but it contains a constant that does not exist. Then — the rounds.
3. Round one: nuclear physics before the pilot¶
The architect: "Which nuclear processes were proven before a pilot, and what did that change in nuclear physics? By your logic all of nuclear physics is nonsense because there was no pilot."
| Year | What | How proven |
|---|---|---|
| 1933 | Szilard conceives the chain reaction | a thought, a patent; Rutherford the same year calls nuclear energy "moonshine" |
| 1938 | fission of uranium | measured by Hahn and Strassmann, explained by Meitner and Frisch |
| 1939 | 2–3 neutrons per fission | measured by Joliot-Curie and by Fermi independently; before that the chain reaction was a hypothesis for six years |
| 1939–1942 | capture cross-sections, moderation in graphite, purity of materials | hundreds of measurements; Fermi built "exponential" piles that never reached criticality, solely to measure the coefficient |
| 2 December 1942 | the Chicago pile, k = 1.0006 | half a watt of power, zero use, one purpose: to check that constants measured separately give together what was calculated |
Nuclear physics was not nonsense before the reactor. But Szilard from 1933 to 1939 was exactly in the protocol's position by the first version: the equation correct, the key constant missing. The number of neutrons per fission cannot be obtained by calculation — it is a property of the nucleus, it can only be measured; until it was measured, Rutherford was entitled to his "moonshine." They measured it — and in three years reached the pile. And still at Trinity in 1945 the physicists placed bets from zero to 45 kilotons, and Bethe separately calculated whether the atmosphere would ignite: a calculation with an unmeasured constant is a hypothesis, however beautiful. The round's outcome: the protocol is "at the stage of Szilard 1934."
4. Round two: the mini-experiments already exist¶
The architect: "The pilot did not require exploding a bomb — and that is what you are advising me. Before the explosion there were mini-pilots, but our mini-pilots have long existed everywhere. Which mini-pilot proves AB-EXIT the way the mini-experiments proved fission before the bomb?"
A check by constants, part of the sources already in the repository (019d, 023c, 023f):
| Constant | Analogue in physics | How measured | What is proven |
|---|---|---|---|
| The price of a vote | fission of uranium | field experiments on vote-buying (São Tomé — Vicente; the Philippines — Hicken et al.), "abstention buying" in the 19th-century US (Cox, Kousser), surveys on selling the vote | the vote is sold, the order of the price is known; money was taken precisely for not coming — that is button B |
| Who leaves when participation gets dearer | neutrons per fission | abolition of compulsory voting in the Netherlands (1970) and Venezuela (1993): turnout −20 points, the poor, young and less educated left; rain on election day (Gomez et al., 2007: an inch — about a percent of turnout); off-cycle election dates (Anzia) | the short horizon leaves first; sign and order are known; the magnitude is measured for a price in inconvenience, not in money |
| Does composition change policy | the chain not dying out | Fowler (2013): compulsory voting in Australia shifted the result to Labor and raised pension spending; Bechtel et al.: the canton of Vaud; 023c — "what happens when everyone is herded to the urn" | composition determines policy, and adding the short horizon moves it toward the short horizon; the protocol does the same with the opposite sign |
Only the assembly is unmeasured: the three constants together, in one jurisdiction, with payment from the budget and a counter. Fermi had all the constants by 1941 — and still built the pile, because k > 1 is a property of the configuration, not of the separate numbers. The round's outcome: not "Szilard 1934" but "Fermi 1941": the constants are measured, the pile is not assembled; the dispute with Rutherford is closed.
Plus one observation at the scale of a country, which Fermi did not have: "Trump's $5,000" (059l). A promise without a source or a law, from one party — 48 % for, 40 % against, ages 18–29 — 61 % for: there is demand for money instead of promises, the myth of the sacred vote is dead. The promise from power did not buy the hall: the party lost, the number vanished, as 059l predicted. The direction was the reverse of the protocol's (payment for participation, not for exit), so the share of takers for exit is not measured by it; but "people will be offended" and "people will not take it" cannot be said after September 2026.
5. Round three: every part is tested in economics and business¶
The architect: "Nuclear fission is an ordinary assembly of what had long existed and had long been tested in parts. All the parts of AB-EXIT have long been tested and proven in other fields, especially in economics and business."
| Part of the protocol | Where tested | What is proven |
|---|---|---|
| voluntary exit for a payment that sorts people by price | overbooking since 1978: when seats are oversold, volunteers are paid to give theirs up (Julian Simon's idea, 1968); before that passengers were bumped involuntarily | those leave for whom the seat is worth less than the compensation; involuntary bumping almost disappeared; the price finds itself |
| a buy-out of a share from which the remaining shares grow | share buybacks: the company pays those leaving, each remaining holder's stake grows without a single new vote | the arithmetic of the share, the same as "×3" (1d, statements 11–12) |
| non-voters amplify voters | shareholder and cooperative meetings: the share is counted from votes cast, abstainers cede weight | decades of practice, nobody considers it a violation of rights |
| payment for not using a resource | payments to farmers for not sowing in the US since 1933, EU set-aside, fishing-licence buybacks, carbon credits | people take money for giving up a right, first those to whom the right yields least |
| voluntary redundancy programmes | corporations, universities, armies in downsizing | those take it whose attachment to the place is weakest (caveat — §7) |
| revealed preference instead of stated | microeconomics after Samuelson; Vickrey auctions | intensity of preference is measured only by price; the ballot does not carry it, money does (040bn) |
| a floating rate from the median | indexation of pensions and minimum wages, automatic stabilisers | self-calibration without a manual lever |
| a rule the elected cannot touch | Swiss and Californian referendums, constitutional debt limits | protection from whoever writes the law for himself (040u item 4) |
| the precinct balance | double-entry bookkeeping since the 15th century | convergence as control, not trust (048m art. 10) |
| a ban on pressure on the choice | labour law, whistleblower protection | works where there is a suit with damages |
The closest whole analogue is overbooking: the right for everyone, compensation to volunteers, those leave for whom it is cheaper, those stay for whom it is dearer, coercion disappears, and the one who gains from the freed seat pays. The round's outcome: the seed is assembled from the tested, and one whole analogue has worked for half a century.
6. Round four: the car — the architect's move¶
The architect: "One more analogy, with cars. Yes, an assembled car must be tested, but can all its units be tested separately too? And is that how car makers do it?"
That is exactly how they do it, and their order is strict:
- Units — on benches. The engine on an engine dynamometer (thousands of hours until its life is counted), the gearbox on a bench, the brakes on a brake dynamometer, the suspension on a shaker rig, the electronics on a simulator. They come to assembly with the measured characteristics of every unit.
- Assembly — for calibration. The whole car is tested not to find out whether it drives: that is known. It is tested to tune how the units work together — noise, vibration, matching gearbox to engine, firmware. Not "yes or no," but numbers.
- Homologation — for the regulator. The crash test and certification are mandatory by law even for an engineer who is one hundred percent sure. Not because the regulator thinks the car will fall apart, but because the right to go on the road is issued not by the engineer.
A manufacturer never finds out at the whole-car stage that the engine does not work. He finds out the setting — and receives permission. These are exactly the two questions that remain for the protocol, and neither of them is "will it work":
| Stage at a car maker | For the protocol | The question |
|---|---|---|
| unit benches | §4 (constants), §5 (parts) | closed |
| assembly and calibration | the courtyard pilot (048l) | "how much to sow": what percent of the median gives what share of takers |
| homologation | the referendum | "who allows it": the voter who votes on the rule judges by image, not by formula |
7. Round five: three objections about the assembly and how they were removed¶
After §5 the assistant left three places where the assembly might not behave like its parts. The architect removed each, and the record matters as a template: each objection was transferred from a field where the object has a property the vote does not have.
| The assistant's objection | The architect's move | The check | Outcome |
|---|---|---|---|
| "In voluntary redundancies the most employable leave first — selection may go the wrong way" | "Where do they go from a ×2 vote? That is a different logic — there is no place to exit to" | the employable leaves because there is a better place; exit under the protocol has no "better place" — only money, and the vote burns. The remainder: in a foregone district the vote is worth zero, and the thinking voter will take B rationally — but the district was foregone anyway, and as soon as it turns sick, the thermostat (1h) returns the burned exactly when his vote starts to decide | removed; the remainder is a budget line, not a selection failure |
| "Farmers paid not to sow planted their worst fields and took the money; campaigns adapt: 'don't take the money, go and vote'" | "The farmer loses nothing, the one who exits loses his vote. The doubters are not a conspiracy against anyone — the result simply does not matter to them, like to a passenger when he flies. You palmed a card" | the farmer's fields are divisible, the voter has one decision — nothing to divide. Campaign adaptation is not a threat but the mechanism: today busing is free, under the protocol a campaign must outbid live money (prohibited) or persuade (the third lever, which does not work on the indifferent); this is the "busing −90–95 %" of 030b, recorded as a risk by mistake | removed |
| "Stigma: taking money instead of a vote may be shameful in the first cycle" | "Then not voting would be shameful too, and nobody would claim benefits out of shame" | a third stay home in a presidential year and half in a midterm — without shame; SNAP is claimed by about 80 % of those eligible, the EITC by about 78 % 🟡, with queues and paperwork; the protocol's choice is secret | removed |
| "The payer is not the winner: the airline pays and itself gets the seat; here the budget pays and the remaining get the share" | "The remaining are at least two, and both get an equal share — that is fair" | full symmetry: everyone is a payer through taxes and a chooser; nobody receives what any other could not; the airline has no such symmetry | removed; fairer than overbooking |
| "The fence: courts, the federal vote-buying statute — it is tested not by a pilot but by the first lawsuit" | "Fences were built by people, and people will tear them down; 'not allowed, then allowed' — it is in the repo" | compulsory voting was introduced and abolished, the franchise was extended five times, Citizens United rewrote in one decision what counts as buying influence; "not allowed" in electoral law is a date, not a law of nature | removed; the fence decides when sowing happens, not whether it grows |
8. Outcome: the seed, the field and the seeding rate¶
The architect: "I see the demand for a pilot as important — but not key. The protocol invents nothing new; it carries an old seed to a new field: the GPS coordinates differ, but the soil, the climate and everything else are like the place where this seed has yielded results for years. The pilot matters — but it is unlikely that everything turns 180 degrees and the seed never grows here."
Accepted, with a clarification of the pilot's role, which is the answer to everyone who says "show me the pilot":
- The pilot does not answer "will it grow." That question is closed by the neighbouring fields (§4–5) and the unit benches (§6). Whoever demands a pilot for that question demands a crash test to learn whether the wheels turn.
- The pilot answers "how much to sow." What percent of the median gives what share of takers at a given stigma, income and set of campaigns. A seeding rate, not a germination test — and it differs between Missouri and Chișinău, because a neutron has one cross-section while a voter is a parameter with feedback.
- The pilot is key for adoption, not for truth. Oregon 2024: a measure paying everyone lost 77 to 23 without a single example that it works. A referendum is won not by a calculation but by the line "in such-and-such district so many took it, turnout of the remaining was such, the mayor changed," because the rule is voted on by the voter who judges by image (1f, 059o §4). Half a watt from Chicago convinces him more than a thousand cross-sections. That is homologation, and the engineer does not issue it.
Where the protocol stands. The seed is tested part by part; the closest whole analogue has worked for half a century; the phenomenon was observed at the scale of a country in September 2026. Not one objection about the soil survived a check of the property it rested on: no "better place," no divisibility, no stigma stronger than benefits, no payer asymmetry, no eternal fence. Two numbers remain that cannot be derived: the seeding rate (p and the share of takers at it) and the date on which sowing is allowed. The first is measured by a test strip the size of a courtyard (048l); the second is decided by the people who built the fence.
Summary¶
| What they will say | What to answer |
|---|---|
| "No pilot — not proven" | All parts (§5) and all constants (§4) are proven; the assembly is unmeasured, and it is measured not by "a pilot in general" but by a test strip the size of a courtyard, for the seeding rate, not for germination |
| "People are not formulas" | People are measured: the price of a vote, who leaves when it gets dearer, how composition changes policy — field experiments and natural experiments in five countries |
| "This is new, there are no analogues" | Overbooking has worked whole for half a century; the other eight parts — from double entry to indexation — are older than anyone in the dispute |
| "Nuclear physics was tested first too" | Tested part by part, in three years, and the pile gave half a watt; the protocol is at Fermi 1941, not Szilard 1934 |
| "A car is still tested whole" | Yes — for calibration and homologation, not for "will it drive"; nobody learns of a dead engine at the whole-car stage |
| "What if the wrong people leave / campaigns adapt / it is shameful / the wrong party pays / the courts forbid it" | Five objections, five checks, all removed (§7): the vote lacks the property each analogy rests on |
Weak spot¶
The chapter records a dispute in which one side is the assistant, and his surrender of positions may look like compliance rather than checking. The defence against that is the table in §7: each row names the property the analogy rested on and shows that the vote does not have it; the reader can check that property himself. The architect's move on this caveat (5 November 2026): "Yes, the assistant is not a person, but logic belongs to nobody — it is either there or it is not. And the arguments will be checked by those who need to, and they are unlikely to be weaker than what half of humanity would offer." Accepted: the question is not who yielded but whether each row of §7 stands under checking; whether the checker is a person or a model has no bearing on the result of the check. The figures on benefits, turnout and rain effects are orders of magnitude from memory 🟡, not checked against primary sources in this conversation. "Fermi 1941" is a metaphor for the position, not an estimate of the probability of success: Fermi had one k and one configuration, the protocol has a parameter with feedback that differs in every jurisdiction, and more than one test strip is needed. And the main thing that stays true whatever the outcome of the dispute: the share of takers at a given p has been measured nowhere, and the register 1i is right that "no pilot has measured it."
Related: 059o §7 · 1g "Adopted nowhere, no data" · 1i · 1h "thermostat" · 048l · 059l · 059g · 019d · 023c · 023f · 030b · 040bn · 040m