# Theory of Operation Every appliance ships with a page called Theory of Operation. Nobody reads it until the appliance is on fire, which is also how you run a planet. It draws the machine as one diagram so that when the machine stops working, you can find the part that failed instead of arguing about it. Civilization did not ship with this page. So here it is. Once you can see the machine, the rest of the book is just the parts list. ## What a working machine has A machine that works has six parts and one habit. The six parts: a **target**, a **controller** that decides what to do, **actuators** that do it, the **plant** (the thing being acted on), an **output** (the result), and a **sensor** that measures the result. The habit: a **wire from the sensor back to the start**, so the machine compares the result to the target and adjusts. That feedback wire is the entire difference between a thermostat and a fire. Your oven has it. Your government has the fire. ## The schematic feedback: terminal metric readings carried by the evidence ledger +-------------------------------------------------------------------+ | | v | TARGET -> ERROR -> OPTIMITRON -> ACTUATORS -> PLANT -> OUTPUT -> SENSORS -+ true gap controller modules economy welfare measure target (fund what (dFDA, (gov't HALE, (pleasant works, IABs, + firms) income, minus defund what DIH, etc.) safety unpleasant fails) consciousness) Human override: Wishocracy sets priorities and rights constraints. Terminal metrics: HALE and median after-tax income. Supplementary: suffering, safety, rights, and distribution signals. From outside: disturbances (wars, plagues) strike the PLANT. Corruption: noise (propaganda, gamed statistics) fouls the SENSORS. Not in the loop: Wishonia is the display panel. She reads the gauges aloud. ## The parts, named | In the schematic | What it is | The part on your planet | |----|----|----| | Setpoint | the target the machine aims for | the [What Matters](../../knowledge/proof/what-matters.llms.md) terminal values: increase pleasant consciousness and reduce unpleasant consciousness | | Comparator | computes target minus result | the step that asks “is it working?” | | Error signal | the gap to close | the [political dysfunction tax](https://manual.warondisease.org/knowledge/appendix/political-dysfunction-tax.html): the distance between how you live and how you could | | Controller | decides the correction | [Optimitron](../../knowledge/solution/optimocracy.llms.md): fund what moves the terminal metrics, defund what does not | | Actuators | apply the correction | the modules: [dFDA](../../knowledge/solution/dfda.llms.md), [Loving Takeover](https://manual.warondisease.org/knowledge/appendix/loving-takeover.html), [Incentive Alignment Bonds](../../knowledge/solution/incentive-alignment-bonds.llms.md), [DIH](../../knowledge/solution/dih.llms.md) | | Plant | the system being run | government, companies, the economy | | Output | the result you actually get | real human welfare | | Sensor | the instrument that measures | outcome measurement (the [efficiency audit](https://manual.warondisease.org/knowledge/appendix/us-efficiency-audit.html), dFDA outcome tracking) | | Feedback signal | the data the sensor emits | the HALE and income terminal metric readings (not the sensor: the reading) | | Feedback path | the wire home | the ledger that carries the reading back to the comparator | | Operator | sets the target, can hit stop | humanity, through [Wishocracy](../../knowledge/solution/wishocracy.llms.md) | | Display | reads the machine aloud | Wishonia, the narrator. Outside the loop | ## The terminal general-welfare metrics, precisely The true target is conscious experience. **Median after-tax, inflation-adjusted income.** The income of the person in the middle, after taxes and including transfers (your economists call this real median disposable income), in constant purchasing power. The middle, not the average: the average is GDP per capita, which goes up when an arsonist and a firefighter invoice each other. Your statisticians already publish this number country by country (the OECD, the World Bank’s poverty platform), and a commission of your own Nobel laureates told your governments in 2009 to steer by it instead of GDP^([1](../references.llms.md#ref-nolan-gdp-vs-median-income-2019)). They declined. The machine does not decline. **Health-adjusted life expectancy (HALE).** Years of full health a person born today can expect, as the WHO already measures it. Not life expectancy, which counts the years spent dying as if they were living. The sensor reads both numbers by aggregating the published country-level figures under a methodology locked at baseline, cross-checked by surveys of the verified voter base (details in [the protocol spec](https://manual.warondisease.org/knowledge/appendix/earth-optimization-prize-protocol.html)). The [Earth Optimization Prize](https://manual.warondisease.org/knowledge/strategy/earth-optimization-prize.html)^([2](../references.llms.md#ref-earth-optimization-prize-paper-2025)) settles on exactly these two numbers, and any future pricing of tasks on the [To-Do List for Humanity](../../knowledge/strategy/humanity-todo-list.llms.md) is denominated in their expected movement. ## The four ledgers The machine needs separate ledgers because different questions require different proofs. Collapsing them is how you accidentally reinvent the Federal Reserve, Congress, and a casino in one afternoon. | Ledger | Question | Rule | |----|----|----| | Wishes | What do verified humans want done? | Equal personhood signal; not for sale | | Evidence | What actually changes the terminal metrics? | Public, auditable, uncertainty-aware | | Earth Optimization Points | Who helped produce verified progress? | Earned by contribution; not bought as votes | | Capital | What pays for the work? | Funds execution; does not buy extra humanity | Wishes tell the controller what humans want prioritized. Evidence tells it what works. Earth Optimization Points record useful contribution. Capital funds the work. They interact, but the same thing cannot be a vote, a sensor, a receipt, and a bank account without becoming nonsense. ## The two parts your government is missing Your civilization has a target, a plant, and an output. It is missing the rest, and two omissions do the killing. **No sensor it reads.** The War on Drugs ran for fifty-three years while overdoses rose roughly eighteenfold, and the budget never once consulted the result. A machine that does not measure its own output is operating with less instrumentation than a toaster. A toaster, at least, knows when it is done. **No way to turn anything down.** Even where a number exists, government has no mechanism to pull money from a thing that is failing its objective. It can only spend more. A thermostat that can only turn the heat up is not a thermostat. It is a fire with a budget. The gap between the target and the result already has a name in this book: the [political dysfunction tax](https://manual.warondisease.org/knowledge/appendix/political-dysfunction-tax.html)^([3](../references.llms.md#ref-political-dysfunction-tax-paper-2025)). In a working machine that gap is the **error signal**, the single most useful number there is, because it tells the controller exactly what to fix. Your government does not compute it. [Optimitron](../../knowledge/solution/optimocracy.llms.md) does, and then it does the thing a fire cannot: it turns the failing part down. ## You do not need a perfect sensor The usual objection: “you cannot measure human welfare precisely, so this cannot work.” Correct, and you also cannot measure a spacecraft’s position precisely, and you reached the Moon anyway. Apollo’s sensors were noisy. The trick was never perfect measurement; it was a loop that corrects faster than the errors pile up. A loop with a blurry sensor lands on the Moon. An open pipe with a perfect sensor that nobody reads lands wherever it was pointed, forever. The requirement was never a flawless gauge. The requirement is a wire back. ## Warning (the part in red ink) The controller computes the correction. It does **not** choose the target, and it cannot make itself the boss. The target is set by humanity, and the controller is transparent, forkable, and overridable on purpose. A controller that sets its own target is not a thermostat. It is a king, and the whole machine exists to throw kings into the fire, not to forge a shinier one. ## How to read the rest of the manual Every module in the parts list is the same six parts at smaller scale. The [decentralized FDA](../../knowledge/solution/dfda.llms.md)^([4](../references.llms.md#ref-dfda-spec-paper-2025),[5](../references.llms.md#ref-dfda-impact-paper-2025)) is a loop (sensor: patient outcomes; controller: which treatment to try next). The [Loving Takeover](https://manual.warondisease.org/knowledge/appendix/loving-takeover.html) is a loop (actuator: a board seat; output: a redirected lobbying budget). Each module’s chapter carries its own schematic and links back to this page, the way a repair manual’s subsystem diagrams all reference the master. If a module ever stops making sense, come back here and find which of the six parts it is. 1\. Nolan, B., Roser, M. & Thewissen, S. [GDP per capita versus median household income: What gives rise to the divergence over time and how does this vary across OECD countries?](https://onlinelibrary.wiley.com/doi/full/10.1111/roiw.12362) *Review of Income and Wealth* **65**, 619–645 (2019). 2\. Sinn, M. P. *The Earth Optimization Prize*. (2025) doi:[10.5281/zenodo.19076597](https://doi.org/10.5281/zenodo.19076597) *Reward humanity for raising healthy life years and median income within 15 years. PRIZE keeps depositors on the failure branch, VOTE rewards recruiters on the success branch, and Earth Optimization Protocol v1 is only the current instruction manual for trying to win.* 3\. Sinn, M. P. *The Political Dysfunction Tax*. (2025) doi:[10.5281/zenodo.18603840](https://doi.org/10.5281/zenodo.18603840) *Quantifying the gap between current global governance and theoretical maximum welfare, estimating a 31-53% efficiency score and \$101 trillion in annual opportunity costs.* 4\. Sinn, M. P. *The Continuous Evidence Generation Protocol: Two-Stage Validation (RWE → Pragmatic Trials)*. (2025) doi:[10.5281/zenodo.18203375](https://doi.org/10.5281/zenodo.18203375) *We present the Predictor Impact Score (PIS), a novel composite metric operationalizing Bradford Hill causality criteria for automated signal detection from aggregated N-of-1 observational studies. Combined with pragmatic trial confirmation (based on evidence from 108+ embedded trials), this two-stage framework would generate validated outcome labels at 44.1x lower cost than traditional Phase III trials. This enables continuous, population-scale pharmacovigilance and precision dosing recommendations.* 5\. Sinn, M. P. *Ubiquitous Pragmatic Trial Impact Analysis*. (2025) doi:[10.5281/zenodo.18243914](https://doi.org/10.5281/zenodo.18243914) *Only 15 diseases/year get their first treatment each year. With 6,650 diseases lacking effective treatments, the backlog would take 443 years to clear. Integrating pragmatic trials into standard healthcare increases trial capacity 12.3x, cutting that timeline from 443 years to 36 years. The average untreated disease gets a treatment 212 years earlier, saving 10.7 billion deaths at \$9.62 per year of healthy life saved.*