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.
59o. The Ship That Turns Around Every 700 Days: the 2026 Midterms, the Hall Instead of the Country, and What of the Protocol Is Already Tested¶
Chapter: 10 — The USA File: 10_059o · v1 · 5 November 2026 Source: a conversation between the architect and the assistant on 4–5 November 2026, the day after the midterm elections of 3 November. It began with a private question (two Dan Sullivans on the Alaska ballot), moved to the question "is it normal that in 700 days the voter turned 180 degrees," then to an analysis of who comes to vote and how the protocol changes the composition of the hall, and ended in a dispute about what in the protocol is proven and what a pilot measures (the dispute is moved out to 040bt). Five corrections by the architect are recorded as moves; all election results are projections by the networks and the Alaska Division of Elections as of the morning of 4 November, not certified totals. A continuation of 059g (the Missouri scenario), 059l (the $5,000 as a rehearsal) and 059n (the pendulum and the third candidate).
How to read. The architect's main thesis is not about parties. It is that a country of such potential and power lurches left and right with no visible cause and loses on every turn not a year but its horizon. The chapter checks the thesis against the figures of 2024 and 2026, breaks the hall into the states of the glossary (1h), shows what exactly the protocol changes, and gives a pointer to the dispute about the pilot (040bt), where the assistant's path from "proven by calculation, a pilot is needed" to "the seed is tested, the field is similar, the pilot measures the seeding rate" is recorded. All percentages for turnout and shifts are estimates 🟡 except where a source is named.
1. The architect's thesis¶
"Trump won crushingly two years ago, and now his party lost just as crushingly. Is it normal that in 700 days the voter changed his principles by 180 degrees? Nothing especially strange — no war, famine, epidemic, tsunami, drought — happened in the country. Perhaps an electoral system that gives such different results in two years without cataclysms and a million dead needs a patch?"
And the conclusion, said at the end of the conversation: "I find it strange that such a big and important ship spends so much strength and energy on 180-degree turns every 700 days. This is expensive for the world and for humanity. And it is strange to me that it happens without visible important causes. If AB-EXIT can fix this, I consider it important as the architect."
The chapter is a check of this thesis. The short answer: the country does not lurch, the hall does; the turn is real, it has no cause, and the protocol strikes exactly the amplifier that turns a shift of a few points into a picture of a rout.
2. What happened on 3 November 2026 — the figures¶
After 2024. Trump — 49.8 % against Harris's 48.3 %, a margin of 1.5 points, one of the most modest victories in decades; the Electoral College 312 to 226 — seven swing states went one way by 1–3 %. The House 220 to 215, the Senate 53 to 47.
Projections as of the morning of 4 November 2026. Democrats: the House — at least 250 seats with 218 needed for control (estimates up to 258–268); the Senate — at least 54 with two races undecided (Alaska, Mississippi). Seven Senate seats flipped, all in states Trump won in 2024:
| State | Trump in 2024 | Lost | Won |
|---|---|---|---|
| Texas | +14 | Ken Paxton | James Talarico |
| Ohio | +11 | Jon Husted (appointee) | Sherrod Brown |
| Iowa | +13 | Ashley Hinson | Josh Turek |
| Kansas | +16 | Roger Marshall | a Democrat |
| South Carolina | +18 | Lindsey Graham | a Democrat |
| North Carolina | +3 | Michael Whatley | Roy Cooper |
| Maine | Harris +7 | Susan Collins | Troy Jackson |
Governors: Iowa — Rob Sand, 55.5 % to 44.5 %, the first Democrat since 2006; Ohio — Amy Acton, 52.3 % to 44.4 %, beating Vivek Ramaswamy; Georgia — Keisha Lance Bottoms, 54.3 % to 45.7 %, the first Democrat since 1998.
Alaska, the private case the conversation began with. Two Dan Sullivans on the ballot: the incumbent senator Dan S. Sullivan (Republican) and Daniel J. Sullivan Jr., a retired teacher from Petersburg, unaffiliated; the Division of Elections removed him from the primary, the Alaska Supreme Court restored him in June. Preliminary: Peltola 48.0 %, Senator Sullivan 47.7 % (a gap of about 600 votes), the second Sullivan 1.7 % (3,583 votes). In the ranked-choice count on 18 November his ballots go to the second preferences — and at such a gap they decide the outcome. The name trick does not "steal" votes from the senator: whoever confused them but ranked the senator second returns the vote to him; only ballots with no second choice burn. For the chapter it is an illustration: a spoiler (030b, the trick "spoilers −40 %") works exactly on the voter who decides in the booth by the first line.
3. The turn is real. There is no cause¶
Checking the thesis "without visible causes" against the list: no war on the territory, no famine, no epidemic, no catastrophe. There are prices, tariffs and the war with Iran, which Trump himself named as one of the causes of defeat. These are causes for a shift of a few points — and that is exactly how much opinion shifted: by estimate 🟡 5–7 points of the national vote. The rest was done by three amplifiers that have nothing to do with people's principles:
- Single-member districts. A 3 % shift flips 30–40 seats, because the fate of the majority is decided in districts with margins of fractions of a percent.
- The Senate map. In 2026 Republicans defended 20 seats against 13 for Democrats, several in states Trump won narrowly.
- Turnout. The main amplifier, and the only one the protocol targets (§5).
The architect's move: "People cannot think in two years that white became black and black became white. The current elections are an illogical system." The assistant's clarification, accepted by the architect: the system is logical inside itself — ballots are counted correctly; what is illogical is its premise, that the sum of today's moods of those who happened to come equals the will of the country. It is an error of measurement, not of arithmetic. An election is an instrument; if on the same object two years apart the instrument gives opposite readings, the instrument is broken, not the object. What it actually measures — who came, whom the primaries nominated and which ad ran in the final week.
Signal and noise in two states. Maine: turnout 🟡 ≈ 74 % in 2024 and ≈ 62 % in 2026, almost the same hall, and it said "Collins, enough" — that is a signal. Texas: turnout 🟡 ≈ 52 % and ≈ 40 %; the 2024 hall was a third brought-in and rarely voting people, the 2026 hall was not; Trump's +14 and Paxton's loss are not a change of mind but a change of hall. The difference between the substantive part of the shift and the instrument's reading is the measure of how much the instrument lies.
3b. Synchronicity: a scale with a shifted zero and a ship turning on the spot¶
Added 5 November 2026 from the analysis of the election results.
The observation. A commentator on the results: the Democrats won wherever it was even hypothetically possible, and where they lost they improved their result "sometimes by tens of points"; "regardless of the quality of the candidates, whether Paxton in Texas or the quite respectable Collins in Maine, regardless of the party colour of the state, regardless of how Trump did there two years ago, regardless of any electoral factors, the whole country shifted synchronously." The author's explanation: the president turned the campaign into a referendum on himself, the candidates were his "avatars on the ballot".
What follows — more than the author drew.
- Synchronicity is a mark of instrument error, not of signal. Had people changed their minds, they would have done so differently: interests in Texas, Maine and Florida differ. An identical shift everywhere, independent of candidate and of state, is a property not of what is measured but of the measuring device: a scale with a shifted zero is wrong by the same amount under any load. Four "regardless" in a row are a description of a systematic error.
- "Tens of points" cannot be a change of opinion. Opinion moved by 🟡 5–7 points (§3); nobody changes convictions by twenty points in 700 days without a catastrophe. The rest is different people in the hall: the brought-in of 2024 stayed home, the angry came. Synchronicity follows from the same thing: one variable changes the composition for the whole country — a presidential year against a midterm.
- The quality of the candidate is switched off. The odious and the respectable lost alike; in 2024 the strong and the weak won alike. The system does not tell people apart — so a party has no reason to field the strong one; the loyal one is enough (059p §4).
- The ballot is compressed to one bit. 470 races reduced to the question "Trump: yes or no". The voter did not choose the other party — it has the worst rating in the history of NBC polling, 29 % against 52 %, with a lead of +7 — he was dismissing the president with the only button available, and 470 people were punished for one man.
The architect's move (5 November 2026). "The protocol removes the voters who run uncertainly from side to side. After that the candidates stabilise too, and the ship need not travel back and forth every two years and turn on the spot. Better not to sail anywhere at all if for two years you burned fuel and abused the engine and after two years came back to the same place. Yes, that is better than sailing backwards, but worse than sailing forward gradually. Trump and the present situation in the US, in my view, clearly show that the system is thoroughly unstable and inefficient: it stands still instead of moving forward."
What this adds. A shifted zero is not only a measurement error, it is an expense. For two years the country goes full ahead in one direction — tariffs, a war, redistricting, pardons — then the hall changes, and the next two years go on the return. The sum of the two voyages is zero distance at double the fuel and with wear on the engine: institutions, alliances, trust (040bs §7: "we will never be the trusted partner again"). Three states of the ship, in order: sailing backwards is worst; turning on the spot is the present one; sailing forward gradually is what a stable hall gives. Under the protocol the voter without a line of his own, led every two years by a new fear or a new package, takes the payment and leaves by himself; those remain whose position holds longer than a cycle (1f §6b: the burned are hard to persuade, but once they change their minds they do not come back). The hall stops jumping — and the candidates stop adjusting to the jump: the winner is not the loudest in a wave year but the one whose line the hall remembers. The course changes by the 5–7 points that really moved in people's heads, not by a hundred and eighty degrees.
4. Who comes to vote and with what logic¶
| State (1h) | Horizon | Motivation to come | What it does to the result |
|---|---|---|---|
| The "core" (dependent on the budget, older) | to the next payment | "so they don't take it away" | votes as told, stably, but for whoever pays today |
| The "quiet one" | to tomorrow | fear of losing the little there is | follows strength, changes side with it |
| The "apathetic" | zero | only if brought | brought in the presidential year, not in the midterm: the main source of the pendulum |
| The "brought-in" | zero | a bus, an employer, a promise | give the "crushing" victory that is gone two years later |
| The "layer" | a week | an ad, a scandal, a face | decides the outcome when two cores cancel each other; judges by image |
| The "burned" | beyond the cycle | do not come: they know the core cannot be outvoted (1f §6) | produce most of the product and do not take part at all |
Hence the pendulum: in 2024 the hall was two-thirds apathetic, brought-in and layer; in 2026 — the core and the quiet ones, who now fear someone else. None of these groups counts beyond the cycle, so the result reflects the mood of the week. Those whose horizon is long and who pay for all of it are at home.
A check by the president's register. The architect: "How many presidents before Trump talked about someone's ass? This is the humour of children aged 6–10, and the president sends it to the voter because he knows it works. What kind of voter is this?" The count of public, deliberate cases across 45 presidents — one: Obama in 2010, "whose ass to kick" about BP, which turned into a week of criticism; George W. Bush (2000) and Biden (2010) — a microphone caught by accident. For Trump it is the permanent register of the stage, and in the 2026 campaign he described it outright as a tool: 3 October, Vandalia, Ohio — "I looked at Husted's ass, and I've never done that with a man before"; 25 October, Port Huron, Michigan — he praised the behind of Nicki Minaj, who was neither on the ballot nor in the hall, and added to candidate Mike Rogers: "I'd say that about your ass too, if it gets you a guaranteed victory." The register's addressees are the layer (vulgarity reads as "real, not of the elite"), the apathetic and the brought-in (a show holds attention for two hours), the quiet one (a signal of strength: whoever can humiliate can protect). On whom it does not work — the burned: 1f §6, fastidiousness about the product, "someone who hires people himself will not spend an evening choosing between two he would not hire." Husted lost to Brown; the White House named "candidate quality" as the cause. A signal tuned to one hall landed in another — and that is observable.
5. What the protocol changes — and why districts stop being an amplifier¶
The assistant's first version was pre-protocol: the protocol as a turnout filter under which the district map and the Senate calendar work as before, and "the protocol does not fix this." The architect's move: "Reread the repo. AB-EXIT changes everything: it removes the whole base and brings out the burned. You again left the logic for a pre-AB-EXIT analysis." A check against 1f, 1h, 1i, 030b, 059k confirmed the correction.
The protocol does not filter the core, it replaces it: the dependent, the apathetic, the quiet and the brought-in take B; the burned return, because their "neither one nor the other" is counted and paid for the first time. It is a different hall, not a thinned old one. Hence:
- Districting loses its target. Gerrymandering works only with a predictable composition: the map is cut to a core that always comes. 030b, trick 7: "districts can be cut, composition cannot, because the composition selects itself." A map cut to a core that has left cuts air. Honestly and by the repo: 030b puts −50 % on districting, not −100 %; the other tricks (busing −90–95 %, turnout suppression, spoilers) are cut harder.
- Primaries change first. Today the candidate is chosen by 15–20 % of a radical core (059k); Paxton, Jackson, Miller came from there. In the same hall with the burned, a candidate with a trail does not reach November; half of the 2026 flips are a punishment for unfitness, not a change of course (059g §3, "fitness has no buyer").
- The same hall in both years. Whoever chooses A chooses A in the presidential and in the midterm year — he has a stake beyond the cycle. The turnout amplifier disappears not because turnout evened out but because those who remain do not depend on the bus.
- The hall is consistent. 1f §6b: the burned is hard to persuade, but once he has changed his mind he does not return. Such a hall gives Maine's signal, not Texas's noise.
- The counter works every cycle. The share of takers by district shows discontent in real time; change comes at a small deterioration rather than accumulating for two years until an explosion (1i).
- The politician knows this from day one. If a hall with a horizon beyond the cycle judges him, "tighten belts for two years for a result in ten" becomes profitable (059). The horizon of power stretches after the horizon of the voter.
Two states under the protocol 🟡
| Texas | Maine | |
|---|---|---|
| Who takes B | the core of the Rio Grande Valley, brought-in young men, dependent rural pensioners, part of the suburban quiet | hardly anyone: turnout 74 %, the core comes on its own |
| Who returns | entrepreneurs, professionals, moderate Republicans who for 20 years "held their nose or stayed home" (059k) | few: almost the same hall |
| The 2026 primaries | Paxton does not knock out Cornyn | Platner and Jackson do not pass |
| November 2026 | a substantive race between two fit candidates, a gap of a few points; who wins — the repo does not promise | Collins is judged on substance, without a scandal background for her rival |
| The 2024→2026 pendulum | from 20+ points to a few | was substantive anyway |
What the protocol does not promise. Who wins: the band around the median exists in every party and state, the number has no party sign (1i). The assistant's first version, "the protocol would have shaved Trump's 2024 peak," is removed by the same line of the register: the protocol strikes not a party but a way of winning — by mobilisation, fear, busing; whichever of the two relied on it more loses, and the repo does not undertake to say in advance which.
6. What the turn costs — for the US and for small countries¶
Corruption, procurement and ghost projects are counted in percent of GDP per year. The horizon is counted by compound interest over decades — and therefore costs more than all the rest together (019 §59, the fifth layer). Every change of course does not cancel the previous one, it devalues what was built under it: tariffs introduced — cancelled — introduced; subsidies adopted — wound down; climate agreements: in — out — in — out. An investor planning a plant for 15 years sees rules for 2 years — and postpones. Postponement is not visible as a loss, only as the difference between what there is and what there could have been.
| Growth-rate difference | After 30 years | After 50 years |
|---|---|---|
| 1 point a year | GDP a third smaller | 40 % smaller |
| 2 points a year | 1.8 times smaller | 2.7 times smaller |
Argentina and Australia in 1900: the same start, resources, immigration; the difference — one changed course every cycle. Today the income gap is 4–5 times, and nobody stole that money. For the US, estimates by the economic policy uncertainty index 🟡: a jump of the 2008–2011 level — about 1 % of GDP and about two million jobs; the tariff uncertainty of 2018–2019 — minus 1–2 % of investment in a year. That is with a pendulum inside one course; the pendulum 2024→2026→2028, where each time the previous is cancelled, is dearer, because the investor has already learned the rule "believe nothing beyond two years." For a small country — Moldova — it is not a share of growth but a vector: East or West, under which everything is rebuilt; the investor waits "until they decide," that is forever, and people cannot wait and leave — a third of the population abroad is the price of a burned horizon. Against that price the patch is cheap: by the 059g calculation the payment at half taking is about 6 % of a state's own revenue per cycle; a lost point of growth is 1 % of GDP every year, forever.
How others got a long horizon. Korea 1961–1987 — a dictatorship; Singapore — one party; West Germany — occupation, a constitution under Allied supervision, a 5 % threshold; Ireland 1987–2008 — an elite bargain on taxes and wages. The common denominator: the short horizon was removed from the hall — by force, by law or by a deal at the top, free and irreversibly for the voter. The architect's move: "But they had no AB-EXIT either, and reaching the same horizon by honest means, like AB-EXIT, is not what happened there." Correct: the protocol reaches the same composition of the hall voluntarily, with compensation, reversibly by referendum and visibly on the counter — and, unlike the four paths, does not rest on whoever imposed it: Korea after 1987 got the pendulum back, the Irish partnership collapsed in 2008 together with the banks.
7. What is proven and what the pilot measures — a pointer¶
The assistant wrote: "For four countries the horizon is proven by result, for the protocol — so far by calculation." The architect: "And what is wrong with the calculation? Or did you count badly?" Then five rounds: three links; nuclear physics before the pilot (Szilard 1933 → neutrons per fission 1939 → the pile 1942); the protocol's three constants and the mini-experiments by which each is already measured; a map of the protocol's parts across economics and business with overbooking as the whole analogue; the car, whose units are tested on benches while the whole car is tested for calibration and homologation; five objections by the assistant about the assembly, removed by the architect's moves. The dispute is moved to its own chapter by the architect's decision — "everyone who reads the protocol will say this": 040bt.
The outcome of the dispute, in one line: the seed is tested part by part, the field is similar, the phenomenon was observed at the scale of the country in September (059l); the pilot does not answer "will it grow," it measures the seeding rate (what percent of the median gives what share of takers) and is needed for adoption, not for truth — Oregon 2024 lost 77 to 23 without a single example. Two numbers remain that cannot be derived: the seeding rate and the date on which sowing is allowed.
Summary¶
| The architect's question | The answer by the outcome |
|---|---|
| Is it normal that in 700 days the voter turned 180°? | The voter did not turn: the shift of opinion is 🟡 5–7 points; the turn was made by three amplifiers — districts, the Senate map, turnout; turnout is the main one |
| What kind of system gives opposite results without cataclysms? | The instrument is sound in arithmetic and broken in its premise: the sum of moods of those who happened to come ≠ the will of the country |
| Who came to vote and with what motivation? | 2024 — the apathetic, the brought-in, the layer; 2026 — the core and the quiet; the long horizon at home in both years; the president's register ("Husted's ass," "if it gets you a guaranteed victory") is a direct admission of whom the signal is for |
| What does the protocol change? | It does not filter the hall, it replaces it; districts and the map remain but stop being an amplifier (030b, −50 %); primaries change first; the hall is the same in both years and consistent; the counter shows discontent before the explosion |
| Is this important for countries or a trifle? | The main thing: the horizon is compound interest; the difference between Argentina and Australia has been counted; for a small country the turn is a vector, and its price is a third of the population abroad |
| What is a calculation worse than a result in? | In nothing arithmetically; the constants are measured by mini-experiments separately, the parts are tested in economics and business; the assembly — the seeding rate — is not measured. The pilot matters for adoption, not for truth (040bt) |
Weak spot¶
All election results are projections as of the morning of 4 November; the Alaska ranked-choice count on 18 November and certification on 25 November may change the figures. Turnout for Texas and Maine, the shift of the national vote, the uncertainty-index estimates and benefit take-up rates are given from the assistant's memory as orders of magnitude 🟡, not checked against primary sources in this conversation. Rally quotes are by press accounts (CNN, Rolling Stone, Newsweek, CNBC), not by transcript. The "two states under the protocol" analysis is a construction, not a measurement; the direction of leaving and returning by group is for the pilot. The assistant's path through the dispute about the pilot, including the mistakes, is recorded in 040bt; here it is only indicated.
Related: 059p · 040bt · 059g · 059l · 059n · 059k · 059 · 1f §6, §6b · 1h · 1i · 1g · 030b · 019 §59 · 019d · 023c · 023f · 048l