The mechanics, in plain terms.
Including the parts that do not work the way you might assume, and the things we deliberately did not build.
Proving a person exists without saying who they are.
Verification runs through ITZA Vault. Your documents are checked on your own device. A commissioned notary looks at them once, confirms a real person is present, and keeps nothing. What is recorded afterwards is a single unique code that means one human was verified. Not a name, not a date of birth, not a scan.
That code is what makes one account one person. It is also all we get, which means we could not identify you if we wanted to.
You can stay anonymous and still publish
Verification is required only to take money out. Everything else - writing, publishing, being read, subscribing, highlighting - works without it, permanently.
Verified is not the same as named
You publish under a pen name. You can hold several. The connection between a pen name and a verified person is never public and is not visible to us in any usable form. Writers who need to publish under a pseudonym can do so and still get paid.
What the composition record contains.
The editor watches a document take shape. When you publish, it signs a summary with your key and timestamps it, so the record provably existed before the piece went out.
Kept, under your key
- The sequence of changes to the document, at one second resolution
- Where text was pasted in, how much, and when
- Session boundaries and checkpoints, enough to replay how it came together
Recorded publicly, in summary form only
- How many sessions, how long they ran, when they started and ended
- Characters typed, characters pasted, largest single paste
- How much was written and then removed
- A signature and a fingerprint of the record, which prove it has not been altered
Destroyed the moment you stop typing
- Every individual keystroke timing
Those never reach a disk or a network. They are used to build the summary and then they are gone. A summary and a signature prove the record is yours and unaltered. A stream of keystroke intervals would be close to a fingerprint, and we do not want it.
The record never acts on its own. It cannot suppress a piece, withhold a payment, flag an account or apply a label. It is shown, and people draw their own conclusions. That is the whole design, and it is deliberate: fast clean typing is what experienced writers do, and any system that punishes it punishes the wrong people.
How payment actually moves.
The platform has its own unit of account. A reader converts into it to pay a writer, and a writer converts out of it when they take money off the platform. Both directions are instant and happen inside ITZA's own exchange, so there is no external market and nothing to trade.
Nothing is ever created as a reward for publishing. Every unit a writer receives was bought by a reader first. That is what makes the arithmetic checkable and what stops volume from being worth anything on its own.
You choose how much to hold
When a payment lands, half converts to the stable unit and half stays as iPress. You set that split yourself and can change it whenever you like. The converted half means a writer earning a modest amount each month has predictable income and is not exposed to a price movement between earning and being paid. The half you keep is a stake in the platform you are helping build.
Subscriptions hold their price
A writer sets what a subscription is worth. Each cycle it is re-quoted so the reader pays the same amount in real terms and the writer receives the same amount in real terms.
Support, and what it actually pays for
Every article carries a support button. A reader can send a writer any amount without a paywall going up or anything changing about what they can read.
For a writer who charges nothing, this is the whole model. Keeping a piece online costs a small amount over time, and support payments are what cover it. New writers start with an allocation that carries them long before anyone has found them, and any piece can be paid off once so it never accrues storage costs again.
No writer is ever billed and no balance can go below zero. If a piece runs out of funding it stops being stored, after four separate warnings and with a one-click export of the work and its composition record. That is the entire consequence, and no judgement about the writing was involved in it.
Nothing is held on our books
Reader money sits in protocol escrow, not in a company account. It releases to the writer as they publish and returns to the reader if they do not.
Why there is a token, and what it does.
You need iPress to pay a writer here. That is what it is for.
Join early and you are given some, whether you came here to write or to read. Writers get more, because they carry the storage costs on what they publish. Spend it however you like - subscribe to someone, pay for a piece, back a writer you think deserves more readers, commission work of your own. Or hold on to it. Nobody is going to make you spend it.
The founding allocation is a fixed amount. It does not get topped up, and nothing here creates more of it for writing, reading, or any other activity. That last part matters more than it sounds. Minting tokens for activity is exactly what killed the platforms that came before this one, because everyone handed free tokens sells them and the price never recovers. It is also why those platforms fill up with machine-written filler: if you pay per unit of activity, the cheapest possible activity becomes the most profitable thing to produce.
So when the allocation is gone, it is gone. From that point, anyone who wants iPress buys it from someone who already has it. The supply does not grow. The demand is however many people want to pay writers here. As the platform grows, we expect that to show up in what iPress is worth.
We are not going to put a number on that or pretend it is a sure thing. If ITZA Press does not attract writers and readers worth paying for, the demand does not appear and iPress is worth very little. There is no buyback, no burn and no supply games to prop it up, and that is on purpose. The token is worth something if the business works. Building the business is the whole plan.
It ends when it runs out, not on a date
The founding allocation is a quantity, not a deadline. Strong uptake uses it up sooner and slow uptake takes longer, and both are fine. We deliberately have not set an end date, because a date is a promise about how fast people will join that nobody can honestly make in advance, and it only creates a scramble to beat the clock and pressure to move the clock. You will always be able to see how much is left.
One ecosystem, not one platform
ITZA, also called ITZACoin, is the hub that connects ordinary money to everything inside. Every platform built on ITZA has its own token, and they all trade against ITZA and against each other on the internal exchange. Tokens earned writing here can be moved to anything else in the ecosystem. None of them are listed on an outside exchange, and none of them need to be.
A nine-page document covering the whole model: why the standard design fails, how the founding distribution works, where demand comes from, and the full list of things we deliberately do not do.
What happens when two people disagree about money.
Commissions, sponsorships, subscriber acquisition and plagiarism claims can all go wrong in ways where somebody is owed something. Those go to arbitration through ITZA's People's Court, with a jury of verified people and the evidence on the record.
Plagiarism is the case where this works better than anywhere else. Two accounts publish the same text. One has a composition record predating publication, the other has a paste. That is not a judgement call, and neither party's follower count enters into it.
What does not go to arbitration
Verification, ranking position and rating weight are not disputes, because nobody decides them. Verification is a published condition of withdrawal. Ranking is a published formula. Rating weight moves mechanically. There is no appeal because there was no ruling, and pretending otherwise would imply somebody could have chosen differently.
Where the writing actually lives.
ITZA Press is not a website hosted somewhere. It is an application that runs on the network itself. Every node carries the application and serves it to readers directly over HTTPS, so the page a reader loads comes from whichever node is closest to them. There is no separate web host and no commercial delivery network in the path.
Articles are stored by their content rather than by their location. Publishing computes a fingerprint of the piece, writes the body into distributed archival storage, and records the fingerprint and a timestamp on the chain. Any copy anywhere in the network can be checked against that fingerprint, so a piece cannot be quietly altered after the fact and cannot be lost because one machine went away.
What that means in practice
- Your work stays reachable for as long as the network runs, not for as long as one company chooses to host it
- Nobody can be pressured into removing it, because there is no single provider to pressure
- Every published version is timestamped and permanent, which is what makes a plagiarism claim provable
- Costs do not scale with a hosting bill, so publishing stays free
You can buy a piece out of storage costs for good
Storage on ITZA normally accrues a small ongoing cost, the way storage does everywhere. The difference here is that you can end it. Pay once for a piece and it is recorded on the chain for good, with nothing left to pay and no payment that can lapse.
That is a real distinction rather than a marketing one. Everything you have ever read online stays online because somebody is still paying for it, and it disappears when they stop, lose interest, or go out of business. A piece bought out of that has no such dependency.
It is optional and always the writer's choice, never automatic and never bundled into something else. It changes nothing except the bill. You decide whether the piece stays published exactly as you did before, and you can unpublish it whenever you want.
To be straight about it, paying once to store something forever is not something only we can do. What is unusual is being able to do it to a piece that also carries a signed record of how it was written and a verified person behind it.
The honest part
Removing a piece removes the body from storage. The fingerprint and timestamp stay on the chain, because that record is what protects every other writer. So delete means unpublish, not erase, and we will say that plainly rather than promise something the design cannot deliver.
The guarantee also depends on people. Content stays available because independent operators run nodes and keep the archive. If you want the guarantee to hold, the most useful thing you can do is run one.
Things we decided against.
- A shared revenue pool. It makes every writer's income depend on a formula, makes each change a redistribution, and gives synthetic writing something to drain. Direct payment removes all three problems at once.
- An applause counter. Anything you can press repeatedly turns into a way of getting attention rather than giving it. Highlights carry more information and cannot be farmed the same way.
- An advertising network. Advertising is one of the reasons writers are leaving the platforms they are on. If a writer sells a sponsorship, the disclosure attaches to the piece and cannot be stripped off.
- Paid placement. There is no way to pay us to appear higher, because we did not build the mechanism. Writers can pay each other to recommend their work, which is between them and visible either way.
- An editorial promotion committee. Human nomination is how you end up with writers trading tips about which editors have promotion powers. Readers decide instead.
- Machine writing detection. It does not work. Detectors miss reworded text almost entirely while flagging honest writers, especially people writing in a second language. We did not build one and we are not going to.
What verification actually stops.
It does not stop a verified person from pasting in machine written text. Nothing could, and any platform claiming otherwise is selling you something.
What it stops is the economics. Low quality synthetic writing works as a business because one operator runs hundreds or thousands of accounts against a shared pot. Take away the pot so money only comes from readers who chose a specific writer, and require one verified person per account, and there is nothing left to scale. That is a smaller claim than the one our competitors make, and unlike theirs it is true.
Read the rest, then come and try it.
Writers are joining first, in small groups. Leave an email and we will let you know when yours opens.
One email when your group opens. Nothing else.