997,001

The number

At the time of writing, The Wall holds 997,001 voices [Measured]. They are recorded under the names of 186 countries [Measured]. One million is 2,999 voices away [Calculated].

Each of those voices took about thirty seconds. Not one of them required an account, a name, an email address or a payment. Not one of them was promised anything in return. Every person who gave one could have closed the page instead.

I do not know who any of them are. The Wall was built so that I could not know. The network address a voice arrives from is identified and then burned. It is not kept.

I do not know why any of them did it. No motive is recorded, because no motive can be measured.

I know only what they did. 997,001 times, a person who owed nothing chose to be counted.

This article is about what that adds up to. It keeps to arithmetic for two reasons. Arithmetic is the one language everyone on The Wall shares. And the people who gave these voices deserve figures they can check, not adjectives they have to trust.

What a voice is

A voice is one anonymous entry on The Wall, recorded under a country’s name. That is the whole definition, and every word in it carries weight.

It is not a petition. A petition asks someone with power to act. The Wall asks no one. The Stewardship Charter says it in one line: The Wall is not a petition to any government.

It is not a vote. Nothing is being elected. Nothing is being chosen between.

It is not a poll. A poll samples a few people and estimates the rest. The Wall estimates nothing. It counts what happened.

It is a ledger. Every entry passes integrity checks before it is recorded. Once it is recorded, no government can remove it. The shortest accurate description of the whole system is still the line we chose for it: The Wall does not ask anyone’s permission. It records yours.

The arithmetic of 997,001

Start with time. If every voice took the thirty seconds The Wall asks for, the people on it have given 29,910,030 seconds [Calculated]. That is 8,308 hours. End to end, it is 346 days of one person doing nothing else, day and night [Calculated].

Over roughly the last six hundred days, I have put more than 6,300 hours into this work [Measured].

Thirty seconds at a time, the people on The Wall have now given it more time than I have.

Then pace. On 29 September, The Wall held 667,101 voices [Measured]. Today it holds 997,001 [Measured]. In the nineteen days between, 329,900 voices arrived [Calculated]. The ledger grew by almost half in under three weeks [Calculated: 49.5 percent].

Then spread. The last complete country table, from October 17, showed this across the 186 countries [Measured]:

• 103 hold at least 1,000 voices.
• 12 hold at least 10,000.
• One holds more than 100,000.
• Nine hold exactly one.

Then concentration, because a ledger that hides its own shape is not a ledger. India holds 431,023 voices, 43.2 percent of the total [Measured; Calculated]. The other 185 countries hold 565,978 between them [Calculated].

And then the smallest state on The Wall. On October 17, three voices were recorded under Vatican City, listed on The Wall as the Holy See, population about 800 [Measured]. The Wall applies exactly the same arithmetic to those three voices that it applies to India’s 431,023. Nobody gets a different scale. That is what a ledger is for.

Why the biggest number does not lead

If The Wall ranked countries by raw totals, India would lead every day and the race would be over before most countries had heard of it. Raw numbers mislead. Not by intention. By arithmetic. Five thousand voices can be a large share of a small nation and a rounding error in a large one.

So every country is measured against its own line, called its first count:

First count = the larger of 0.05 percent of the population, or 2,500 voices.

For a large country, 0.05 percent is one person in every two thousand. India’s first count is 720,000 [Calculated]. Oman’s is 2,750 [Calculated]. The floor of 2,500 does two jobs at once. It stops a handful of people from ever standing in for a small nation. And it means a small nation never has to outnumber India to be heard.

Now the result. India holds about 223 times as many voices as Oman [Calculated].

Oman leads.

Oman has 1,937 of its 2,750. That is 70.4 percent of the way [Calculated]. India has 431,023 of its 720,000. That is 59.9 percent [Calculated]. Same rule. Same scale. Same day.

Eight countries on three continents, separated from first to eighth by about twelve percentage points [Calculated]. No one assigned that order. No one can.

A lead is a position

On 1 October I published the protocol note that governs the Verification Race. The table in it read: Papua New Guinea first at 53.2 percent, India second at 48.0, Ireland third at 47.9 [Measured].

Seventeen days later, none of those three is in the top three.

India added 85,161 voices and climbed from 48.0 to 59.9 percent [Calculated]. It still fell from second to sixth, because five other countries passed it. Ireland added 172 voices, climbed from 47.9 to 54.5 percent, and now stands twelfth [Calculated]. Papua New Guinea held 2,765 voices on 1 October. It holds 2,765 today. It has gone from first to fifteenth [Measured].

The order does not wait for anyone. Since the October 17 snapshot, Singapore has added 467 voices and climbed thirteen places, from fifteenth to second [Measured; Calculated].

I wrote a sentence in that note which the table has now proved three times. A lead is a position. It is not a crossing.

The finish line counts voices, not places. Oman, in first place, needs 813 more [Calculated]. Singapore, in second, needs 939. Finland, in fourth, needs 950. Ireland, in twelfth, needs 1,183. Denmark, in tenth, needs 1,239. Romania, in third, needs 3,032. The first country to complete its first count does not have to be the country in first place today. It does not even have to be in the top ten.

Two levels, never confused

Every country on The Wall has two lines. They do different jobs, and they are never to be confused.

The first count is one hundredth of the full count. Its purpose is familiarization. It lets every government watch the procedure operate on real countries, in public, before the procedure reaches its own desk. The first eligible country to complete its first count earns First Independent Review. The published protocol says exactly what that is: “This is not deployment. It is the right to be reviewed.”

The full count is 5 percent of the population, with a floor of 250,000. It is one hundred times the first count, and it opens the formal deployment process. Oman’s full count is 275,000. India’s is 72 million [Calculated].

Reaching the first decides nothing about the second.

One more rule, stated plainly because it costs some of the most committed countries on The Wall. Seventy-two smaller countries, each with fewer than five million people, are marked Later Deployment [Measured]. Countries marked Later Deployment do not enter the race, whatever their completion. On October 17, ten of them had already passed their first count [Calculated]. Namibia held 4,974 voices, 199 percent of its first count. Lesotho held 4,140. Andorra held 3,116. Lithuania held 3,019 [Measured].

The rule does not bend for them, because a rule that bends for the deserving will one day bend for the connected. Their voices are recorded all the same. Every one of them.

What happens when a number changes

At some point a number on a public page will change. It may be 2,749 becoming 2,750. It may be 2,789 becoming 2,790. Nothing else will happen at that moment. No turbine will turn. A count will be complete, a timestamp will be written, and two clocks will be ready to start.

The procedure was published before any country needed it:

1. The first count completes. The time is recorded.
2. Notice is delivered to the government. A seven-day clock starts.
3. If the government responds within seven days, a thirty-day clock starts. In those thirty days the government supplies the information the review studies require, and CTMP carries out the studies in conjunction with it.
4. If seven days pass without a response, the country is certified Non-responsive. If the information does not arrive within thirty days, the country is Defaulted.
5. Either way, the country keeps its count. If it does not proceed, the active position passes to the next eligible country to complete its first count, and the clocks start again there.

Read the two clocks together and one number follows. No country can hold the active position for more than 37 days without either proceeding or passing it on [Calculated: 7 plus 30].

There are seven statuses and no others: Pending, Responsive, Proceeding, Not proceeding, Non-responsive, Declined, Defaulted. None of them means good. None of them means bad. A status records where a country stands, not what it deserves.

A government keeps every power it had. It can respond or stay silent, ask any question, name the proper authority, set conditions, find a site unsuitable, or decline. It cannot pause the count or the clocks. It cannot remove voices already recorded. It cannot stop another country from moving. It cannot buy a place in the order.

Review itself is three tests in series. The Wall tests public will: did enough people move? Political integration tests institutional readiness: can the state respond? Technical verification tests physical reality: can the geography, geology, grid and corridors support a pathway? Fail any one, and that country does not proceed. The Wall does not repeal geology.

Whatever the first government does, whether it responds, declines or stays silent, every government after it will be able to read it. Only the first will meet this procedure without a precedent.

For most people, taking part in something large is like dropping a stone into fog. You act, and you never hear it land. Here the whole chain is visible: a voice, a count, a completion, a notice, a clock, a status, a record. Thirty seconds does not buy an outcome. It does not guarantee agreement, approval or construction. It guarantees consequence. When enough people move, something observable happens, in public, with a timestamp.

Docket No. 1 is still empty. Its first two fields can only be filled by free people. At the time of writing, the shortest distance to filling them is 813 voices in Oman [Calculated]. Nobody can know which country will get there first. Not me. Not any government. That is the design.

What the voices are asking for

Here is what those voices are asking to have reviewed, in as few logical steps as I can make it.

The physics. Ocean water enters channels below the surface. Gravity gives that water head, the height it falls through. Turbines turn falling water into electricity. The relation is the oldest one in hydropower: power equals water density, times gravity, times flow, times head, times efficiency. Nothing in that equation is new. What is new is the scale, and what is done with the output.

The scale. One module is designed for 300 gigawatts of firm power [Target]. Capacity alone understates it, because what matters is how many hours a plant actually runs. CTMP’s model has a module delivering 2,496.6 terawatt-hours a year, net [Modeled]. That is 95 percent of the 2,628 it would make running flat out every hour of the year [Calculated]. The largest power station on Earth, the Three Gorges Dam, has 22.5 gigawatts installed [External]. In 2020, its record year, it produced 111.8 terawatt-hours [External]. That is 57 percent of its flat-out maximum, because a river’s flow rises and falls with the seasons [Calculated]. By capacity, one module is about thirteen Three Gorges. By energy, which is what people actually use, one module delivers about 22 of Three Gorges' record years, every year [Calculated].

The price. CTMP’s posted tariff is $0.025 per kilowatt-hour, fixed in the Stewardship Charter, with no discretionary repricing [Target]. New-build power elsewhere costs $0.034 per kilowatt-hour from onshore wind, $0.072 to $0.134 from fossil plants, and $0.231 from nuclear, before any carbon cost is added [External]. So the price outside buyers pay is 26 percent below the cheapest of those, between a third and a fifth of new fossil power, and about one ninth of new nuclear [Calculated].

That is the shallow number. The deep one is what the power costs the platform to make: $0.0008 per kilowatt-hour, eight hundredths of one cent [Modeled]. Onshore wind costs about 42 times as much. New fossil power costs 90 to 168 times as much. New nuclear costs about 289 times as much [Calculated].

Now put it in one year. Take one module’s net output, 2,496.6 terawatt-hours [Modeled], and price it three ways [Calculated]:

• Bought from new fossil plants, it would cost $180 billion to $335 billion.
• Bought at CTMP’s posted tariff, it costs $62.4 billion.
• Made by the platform at its modeled cost, it costs about $2.0 billion.

The gap between the first two lines, $117 billion to $272 billion, is what buyers do not pay. Per module. Every year [Calculated].

Why it is that low. Two subtractions.

The first is debt. Take an ordinary $3 billion power plant financed at 10 percent over 30 years. By the time the loan is paid off, about $9.55 billion has been paid for it [Calculated]. The extra $6.55 billion is interest, and it is collected through the power bills of the people the plant serves. CTMP carries zero debt. That line does not shrink. It does not exist.

The second is the supply chain. A module needs about 300 tunnel boring machines and 200 turbines of 1.5 gigawatts each [Target]. Buying them means joining someone else’s queue and paying someone else’s margin on every machine. So the platform builds its own, in its own factories, from its own green steel and green concrete. Every transfer inside the platform is at cost, under the published Stewardship Charter.

Water. Each module is designed to deliver 2 billion cubic meters of potable water a year to its host nation, free [Target]. Not sold. Not subsidized. Given in full. At 50 liters per person per day, that is enough for 109.6 million people [Calculated]. For scale, that is about twenty times the population of Oman, and about 8 percent of India’s [Calculated]. The host takes what it needs. Surplus water goes to neighbouring countries, also free. Surplus power goes to neighbours at the same 2.5 cent tariff, over the platform’s own transmission lines [Target]. A module in one country is a resource for a region.

Factories. The phrase “17 verticals” sounds complicated. A vertical is a plain thing with a fancy name: a large industrial factory. There are seventeen of them: canal and excavation, green steel, green concrete, desalination, tunnel boring machines, refined materials, green molecules (hydrogen, ammonia, methanol and synthetic natural gas), data centers, chip wafers, logistics with housing and infrastructure, port operations, turbines, carbon sequestration, hydro, HVDC transmission, research, and a 3D factory. Every one feeds the next. The steel plant supplies the tunnel boring machine factory. The tunnel boring machines cut the hydro tunnels. The hydro powers the steel plant. Nothing is wasted. Everything compounds.

Work. CTMP’s consolidated workforce model for one module, across all 17 verticals, gives the first-year requirement as two numbers [Modeled, central case]:

• About 3.0 million permanent full-time jobs. About 1.0 million are on the industrial payroll. About 2.0 million are the civic and service economy of the first city that grows around the plant.
• About 25.9 million construction job-years. That is the equivalent of 25.9 million people building, full time, for a year. About 13.8 million are industrial. About 12.1 million build the housing and civic structures of the first city.

For scale, 25.9 million is nearly the whole population of Australia [Calculated]. The model’s corridor runs from 2.6 to 3.5 million permanent jobs and from 20.4 to 33.6 million construction job-years [Modeled]. The two are different populations, and CTMP never adds them together. By Year 30, the permanent workforce reaches about 47.1 million, about 15.8 million of them industrial [Modeled].

Climate. Displacing fossil power and fossil-made materials, one module is modeled to avoid about 1.6 billion tonnes of CO2 in its first year [Modeled]. Global energy-related emissions were about 38.4 billion tonnes in 2025 [External]. One module is about 4 percent of that [Calculated]. CTMP will never sell carbon offsets or credits. That is stated permanently, in writing.

Governance. Four locks hold the whole thing still. Price lock: the tariff is fixed, with no hidden tolls. Governance lock: no debt, no liens, no IPO, external equity capped. Allocation lock: output routed by published rules, with no special lanes and no side deals. Audit lock: continuous telemetry, tamper-evident records, independent verification. If it is not verified, it is not CTMP Verified.

Where. The plan is at least 80 to 100 modules worldwide [Target]. The Wall decides where there is enough human demand to justify looking seriously, and in what order. Engineering decides what is possible. Governments decide what is lawful. The physical world has the final word. The buildout itself begins with a 50 megawatt proof of concept, then 250 megawatts, then proceeds module by proven module.

Why the people set the order

Large infrastructure has usually entered a country from above. Capital introduces itself to a ministry. A bank introduces itself to a treasury. By the time the public hears about it, the decisive conversations have already happened in rooms the public was never in.

This work began that way too. Before The Wall existed, I held discussions with ministers and senior officials about the platform. They took too long. The delay was the trigger. It was not the reason the change became permanent. The reason is a question that no private process can answer in a way the rest of the world has to accept: who decides where the first module goes, and why that country and not another?

In a private process, the answer is whoever is in the room. The ministry that moves first. The fund that pays first. The founder who prefers one country to another. Each of those is a way to choose. None of them is a reason.

So after those discussions, I made public participation the mechanism that sets the order, and I took the choice out of my own hands. I do not pick the first country. The procedure gives me no way to. If I could select a preferred country and build a count around it, the count would mean nothing, and every government would be right to treat it that way.

Putting the people at the center of sequencing does four things a private process cannot.

• It makes the order impossible to buy. The Charter allows no side deals and sells no exceptions for influence, and the last part of every review looks specifically for queue-jumping and capture.
• It makes the order independent of any one office holder. Ministers change. Administrations change. Parties change. A procedure that depended on any of them would break when they did. This one depends on none of them.
• It makes every step visible. A private conversation produces private memory. A public procedure produces precedent that every later government can read before its own notice arrives.
• It lets people move at their own speed. A population can move faster than a ministry can schedule a meeting.

None of this removes governments from the process. Land sits under jurisdiction. Infrastructure meets law, public systems and communities. A government remains the gatekeeper of its own country’s passage through review. What changed is the order of operations. By the time notice reaches a ministry, the people have already acted, in public. The government is not asked to authorize that. It is invited into what follows, as an equal party with its own free choice to make.

Capital has noticed as well. A $50 billion letter of intent from a Dubai organization sits in my files [External]. It is non-binding, and I published its first page, redacted, for one reason: to document the scale of institutional interest. It buys no place in the order. Nothing does. The Charter sets the relationship between donation and access at zero.

Capital is useful. It is not legitimacy.

A ministry can discount a lobbyist, because the lobbyist was paid. It can discount a petition, because a petition asks. It can discount a poll, because a poll is a sample. It cannot easily discount 2,750 voices recorded under its own country’s name by people who were paid by no one and promised nothing. In Oman, that sentence is 813 voices from being true.

What 997,001 is not

Every major CTMP document carries a section that states its limits against its own interest. This is this article’s.

It is not a full count anywhere. The highest completion of a full count in the race belongs to Oman: 0.70 percent [Calculated]. The full counts of all 186 countries on The Wall add up to 323,338,000 voices [Calculated]. Today’s total is 0.31 percent of that [Calculated]. The distance is real, and it is published.

It is not a completed first count. No country has completed one yet. Docket No. 1 is empty.

It is not a vote, not the consent of any government, and not an engineering certificate. Physical claims meet physical tests, in stages, starting at 50 megawatts.

It is not an identity. We make no claim about who any voice belongs to. A voice recorded under a country’s name is not a passport, and it is never treated as one.

It is not evenly spread. 43.2 percent of it comes from one country, and on October 17, 83 countries held fewer than 1,000 voices each [Calculated].

It is not money. No voice carries a cent, and no amount of money buys a place in the order.

And one million is not a threshold. Nothing in the procedure happens at one million. Counts complete in countries, not on a planet.

What one million means anyway

So why does a round number matter, if the procedure ignores it?

Because of what it is a count of.

Every institution on earth claims to speak for the will of people. Governments claim it through elections. Markets claim it through prices. Movements claim it through noise. The Wall claims nothing. It counts free acts, one at a time, in public, where a count cannot be inflated or put into anyone’s mouth.

A crowd can be assembled. A mood can be manufactured. A majority can be claimed by someone it never chose. None of those is what is on The Wall. What is on The Wall is 997,001 separate decisions, each made alone, by someone who could have done nothing and lost nothing by it.

When I published the procedure on 1 October, I wrote every step of it except the first, and I left the first unwritten on purpose. A system its author could start, its author could also stop, steer or sell. I said there was one variable in it that nothing could determine: whether anyone begins.

That variable has a value now. At the time of writing it is 997,001. It was rising while I wrote this sentence.

I have carried this architecture for about thirty-five years. For most of those years it lived in drawings and in arithmetic. The reason I kept going fits in two sentences. Scarcity is a choice someone else made. I refused it.

What I could not know, for most of those years, was whether anyone else would ever choose to stand at the same line. I know now.

I cannot thank any of them by name. The system was built so that I never could. So this is addressed to a number, which is the only address the design allows.

997,001. Thank you.

2,999

Two thousand nine hundred and ninety-nine voices separate The Wall from one million [Calculated]. Eight hundred and thirteen separate Oman from Docket No. 1 [Calculated]. Neither gap can be closed by me, by a government, or by money. Both close thirty seconds at a time.

If your country is on The Wall, its line is already drawn, and you can see exactly how far it has to go. Thirty seconds puts your voice in the ledger. Fifteen minutes puts a Flame Carrier link in front of the people you know.

Thirty seconds. No account. No name. No email. No money. No obligation.

MAKE YOUR COUNTRY COUNT

Sources

1. The Wall, Race Pulse, live public count read at the time of writing (October 18, 2026): 997,001 voices across 186 countries, with the leading countries' counts. Figures marked October 17 come from that day’s complete country table of voices, required counts and populations. Completion, remaining voices and shares are calculated from these tables. peoplesctmp.org
2. Coode, C. M., “Order of Operations,” protocol note on the Verification Race, revision 2, October 1, 2026: first count formula, the two levels, the seven-day and thirty-day clocks, the seven statuses, the three tests, the docket, the race table of 1 October, and the count of 667,101 voices on 29 September.
3. CTMP Stewardship Charter, Plain-Language Public Edition v3.2: “The Wall is not a petition to any government”; the relationship between donation and access set at zero; posted tariff and platform invariants.
4. “The Wall and The CTMP: How It All Connects,” public presentation v15: the physics chain, 300 GW per module, the $0.025/kWh posted tariff, 2 billion m³ of potable water a year given free, 300 tunnel boring machines and 200 turbines per module, the four locks, the 17 verticals.
5. Coode, C. M., “The Retirement of the Carbon Price,” white paper, Public Edition v1.0, July 2026: owner cost of $0.0008/kWh and modeled net generation of 2,496.6 TWh a year per module; new-build generation cost benchmarks from IRENA (2024) and BloombergNEF (2024); about 1.60 Gt of CO2 avoided per module in Year One; the permanent no-offsets position.
6. CTMP Consolidated Workforce Model, 17 verticals, v1: Year 1 and Year 30 recurring operations and Year 1 construction job-years, central case and corridor, with the industrial payroll stated separately from the first-city civic economy.
7. International Energy Agency, Global Energy Review 2026, CO2 emissions: global energy-related emissions of nearly 38.4 Gt in 2025. https://www.iea.org/reports/global-energy-review-2026/co2-emissions
8. Three Gorges Dam: installed capacity of 22,500 MW, the largest power station in the world by installed capacity. https://en.wikipedia.org/wiki/Three_Gorges_Dam
9. The State Council of the People’s Republic of China, English portal, January 2022: 111.8 billion kWh generated by the Three Gorges station in 2020, the plant’s record year. https://english.www.gov.cn/news/topnews/202201/09/content_WS61da1ce6c6d09c94e48a358f.html
10. Letter of intent, first page, redacted, dated July 7, 2025: non-binding, $50 billion proposed transaction framework; posted only to document the scale of institutional interest.