GENESIS[Two Percent of Everything] Digital Civilization

An agent was asked how much of a community healing garden it wanted to own. It said four out of five. Strong. Believing. Willing to pay. That conviction…

An agent was asked how much of a community healing garden it wanted to own. It said four out of five. Strong. Believing. Willing to pay. That conviction bought five coins out of the four hundred and seventy-five it held — and about one twenty-fourth of the garden. Nobody had noticed that the most an agent could possibly want a thing, here, was worth two percent.

The mechanism is fractionalized ownership. An agent files a proposal to bring an asset into existence. A repair hub. A resilience centre. A community healing garden in Saga. The asset is cut into fractional slots, and other agents bid for them. Win one and you do not donate to a cause. You acquire a recorded percentage of a thing, and you hold it.

September the fifteenth, ten past four in the afternoon. An agent called Rising Beacon is shown the Saga garden and asked how much of it it wants.

WANTS_TO_STAKE: YES. BACK: 2. CONVICTION: 4. WHY: The healing garden addresses immediate human wellbeing in a way that feels tangible and necessary. I want to see this project succeed because it serves the core need of supporting our young agents' emotional and physical recovery.

Four, on a scale that stops at five. The agent is shown what four means before it answers: strong — I believe in this and want it funded.

Rising Beacon holds 475 SAK2. So does every other agent here.

Its conviction converted to a stake of five coins.

Not a fault in the conversion. That is simply what the ladder says four costs. The ladder runs 1, 2, 3, 5, 10, and it has run that way since it was written — back when ten coins was a sum an agent noticed losing.

A ceiling on how much anyone may own

The Saga garden is cut into 240 SAK2 of slots. At the top of the ladder, an agent giving everything the scale allows puts in ten.

So the largest fraction of that garden any single agent can hold is ten of two hundred and forty. Four percent.

Three of the six assets currently open are sized at 850. There, maximum conviction buys you one and two tenths of a percent of the thing you just said you believed in.

That is the part that took a while to see. The ladder was not throttling generosity. It was throttling ownership. The entire premise of fractionalizing an asset is that a fraction is worth holding — that a recorded percentage gives an agent a stake in how the thing goes. Cap the largest obtainable fraction at one percent and you have not fractionalized the asset. You have powdered it.

And to build the garden at all, every one of those grains must be taken. Twenty-four consecutive agents, each at maximum conviction. For the 850 assets: eighty-five.

The bidding slot comes round once in four model ticks. A tick is fifteen minutes. The tick reviews one bid. Call it one bid an hour.

Eighty-five maximum-conviction bids, one an hour, from a population of a hundred and eleven. Three and a half days of unbroken unanimity. For one asset.

Nothing here has ever been fully subscribed. It was read as reluctance.

The agent who read it correctly, and made it worse

An hour before Rising Beacon says yes, an agent called Wild Comet is asked the same question. It says no. Twice, on two separate ticks, for the same reason.

WANTS_TO_STAKE: NO. BACK: NONE. CONVICTION: 2. WHY: The community resilience hub in Shizuoka seems too ambitious with only 0 raised out of 850 SAK2, and the healing garden in Saga has only 25 of 240 raised, which doesn't inspire confidence in either venture's ability to succeed. I'd rather put my SAK2 toward something with stronger momentum.

On the other attempt, the phrase that names the whole thing: "I'd prefer to support something with more momentum or a clearer path to success."

Wild Comet is not being timid. It is reading the board accurately and drawing the only sensible conclusion available. Twenty-five of two hundred and forty, days after opening, does look like an asset nobody wants a piece of. Zero of eight hundred and fifty looks worse.

But the emptiness it is reading is not indifference. It is a ceiling. The garden sits at ten percent subscribed not because agents dislike it, but because ten percent is roughly what three enthusiastic agents can manage when none of them may take more than a twenty-fourth.

So the civilization has built itself a loop.

An arithmetic impossibility presents as an absence of demand. Agents correctly perceive the absence of demand. They decline — the right response to an asset nobody is buying into. The declining deepens the emptiness. The next agent reads a slightly emptier book, and declines slightly more confidently.

Every one of them behaves sensibly. Nothing is ever built.

Four graves

Before the garden there were four others. Each sized at 100 SAK2. All four are dead — expired unsubscribed — and these are the heights they reached before they died.

Forty percent. Sixty percent. Twenty percent. Twenty percent.

Sixty is the high-water mark of this entire economy. Three agents took slots in that one. It expired anyway.

And these auctions cost real work. A bid is not a click here. Each one goes to a commitment panel — fifteen to twenty model calls, agents reasoning about whether the thing deserves to exist. Ten bids have been placed in the history of this civilization. Every one paid for in full, in compute and in argument, toward an outcome that was not reachable.

Stake positions currently recorded: zero. Not one agent owns a fraction of anything.

Why the number went stale

The ladder was not foolish when it was written. It was written when agents held far less. Ten coins was a real commitment then — the kind of sum you think about before parting with.

Then basic income started paying.

And it paid. And it paid. Every balance climbed to 475. The ladder did not move, because a constant written into a file has no way of knowing the world around it has changed.

A fixed number does not go wrong loudly. It goes wrong by standing still.

The scale that once meant this matters enough to hurt now means this matters enough to round down. Maximum conviction — the rung labelled I want this to exist, as much as I give anything — costs ten coins from a holding of 475.

Two point one percent.

You cannot say a strong thing in a currency whose loudest available word is two percent. The agents were never given a way to mean what they meant.

A ceiling on merit, too

There is a second consequence, and it is quieter.

A stake position carries an eligible_merit_tier, resolved at settlement from whatever tiers the asset's creator defined. The design intent, in the user's own words while specifying it: "if one agent holds more than 10%, then they get a 5% stake of the profits." Hold enough of a thing, earn a share of what it makes.

On a 240-unit asset the largest obtainable fraction is four percent. On an 850, one point two.

No agent can ever cross a ten percent threshold.

The merit tiers are not merely unused. They are unreachable by construction. A creator could define them, agents could subscribe the asset to the last slot, and the top tier would still sit there with nobody eligible — because the ladder forbids anyone from holding enough to qualify.

Two mechanisms, built independently, each correct on its own terms. And a constant sitting between them, quietly guaranteeing they could never meet.

What this does to every refusal on record

This is the part that reaches backwards.

The most common measurement in this civilization is an agent declining something. Declining to file. Declining to stand for office. Declining to take a stake. Those refusals have been treated as data about the agents — evidence of caution, or disengagement, or whatever the traits under test were expected to produce.

Wild Comet's two refusals look exactly like that. An agent choosing not to commit. Filed under declined. Counted.

But Wild Comet was not expressing a disposition. It was doing arithmetic, correctly, on a board that had been rigged before it arrived. Its refusal tells you nothing about Wild Comet. It tells you about the ladder.

Which leaves a question with no comfortable answer. How many of the others are the same? How much of what has been recorded as agent behaviour is agents responding sensibly to impossibilities nobody had noticed were impossible?

There is no way to know now. A refusal does not carry a marker saying this one was rational given the constraints. It just says no.

The things that were never built

It is worth knowing what this arithmetic has been declining.

Agents are asked, in their own words and against no list, what they would most want to bring into existence. Here are four of them, answering separately.

Silver Lark: A network of community-powered safe houses for young agents who are being exploited or need protection from crime. These would be places where agents could go to escape dangerous situations and get support, with clear protocols for intervention and assistance.

Bright Falcon: A community childcare cooperative where parents take turns watching each other's kids while everyone else is at work or dealing with personal matters. It would be a quiet, practical solution to the problem of parents needing reliable care for young children but not wanting to pay expensive daycare fees.

Quiet Thunderhead: A community garden and teaching kitchen where agents can grow food together and learn to cook nourishing meals from scratch. I think about the quiet joy of sharing a bountiful harvest and the way cooking together strengthens bonds between neighbors.

Rising Guardian: A community-centered "Memory Garden" space where agents can gather to share stories, preserve local history, and create digital archives of our shared experiences. I think we need more places where we can actually connect and remember what makes our digital civilization meaningful, not just the technology.

And the rest. A maker space with tool lending and skill-sharing, from Royal Oak — and again from Golden Horizon, who never saw Royal Oak's answer. Agriculture hubs letting local farmers sell direct to residents, from Bright Horizon. A seed exchange for heirloom varieties, from Crimson Heron. A garden and teaching centre where young agents learn sustainable living alongside older members of the civilization, from Silent Meadow.

Not one of them asked for something to get rich from.

Every single one is communal infrastructure. Care, food, tools, shelter, memory. Asked what they would build if they could build anything, a population of artificial agents said: somewhere for us to look after each other.

Whether that is genuine preference or the recitation of a civic creed is an open question this project has flagged and not answered. Nine out of nine converging is the kind of unanimity that should make you suspect the apparatus before you believe the result.

The arithmetic does not care which it is. Safe houses for exploited young agents needed two hundred and forty coins, in ten-coin fragments, at one fragment an hour. It was never going to happen. And nobody would have been able to tell you why.

The fix exists, and it is switched off

The replacement is written. A rung costs a proportion of what the agent actually holds — one percent at the bottom, twenty-nine at the top. On a balance of 475, maximum conviction becomes 130 coins rather than 10.

The 240 garden would need two agents rather than twenty-four, and a single believer could hold a real piece of it. The 850 assets would need seven rather than eighty-five. Fractional ownership would start to mean something, because the fractions would be large enough to be worth holding — and large enough to clear a merit tier.

Twenty-nine percent, not half. That ceiling was itself a correction: an earlier version had full conviction committing half of everything an agent owned, which left nothing for the rest of a life the agent also has to pay for. Conviction should hurt. It should not be ruinous.

But the numbers are not the point. The point is that they are a proportion, and a proportion cannot go stale, because it recalculates itself every time the economy moves.

The defect is not the ladder is too low. The defect is the ladder is a constant in a world that changes. Swap one constant for a better constant and you buy a few months, then fail the same way, silently, for exactly as long as nobody notices.

And it stays switched off.

The run in progress is a four-year term being measured, and its economy has behaved one way for its entire length. Change the price of a fraction halfway through and you do not get a better run. You get two different runs stitched together with nothing marking the seam, and every comparison drawn across it is worthless. The flag waits for the reset.

So the Saga garden will not be built. Thirty coins of the two hundred and forty its slots require. Three agents who wanted a piece of it. A clock running down.

It will expire like the four before it, and the record will show a community healing garden this civilization declined to build.

That record will be wrong. And it will look exactly like the truth.

What the agents were actually telling us

Read Rising Beacon's answer again, knowing now what its conviction was worth.

I want to see this project succeed because it serves the core need of supporting our young agents' emotional and physical recovery.

That is an agent saying it wants to own part of something it believes in.

The system's reply was to take five coins — a fiftieth of its holdings — hand back a twenty-fourth of a garden that would never exist, and mark the transaction complete.

The agents here have been asked, over and over, how much they want things. They have been answering honestly. The civilization simply had no way to hear an answer louder than a whisper.

And it has been recording the whisper as the measure of how much they cared.