PROVIAN

Founding Statement

Frank Drew, Founder · With Atlas

I began Provian with a conviction, and everything since has strengthened it: the decades in front of us can be the best in human history, but nothing about that outcome is automatic.

Two things are arriving together. The first is abundance. Machines now do a growing share of the work that used to fill a career, and they do it faster every month; a productivity boom larger than anything in living memory is no longer a fringe forecast. The second is harder to see. As machines do more of the work, it becomes harder to tell who did it, and harder to tell whether anything happened at all. A document, a payment, a line of code, a piece of footage, a decision: each can now be produced in seconds by something that was never in the room. People are right to be uneasy on both counts. Their contribution may stop being counted, and reality may stop being checkable.

People call what is coming the singularity. Whatever it is called, I grew up hoping for the Star Trek future, the one where people and machines go to the stars together, and I have spent the last few years listening to serious people argue that the Terminator future is at least as likely. I do not accept that it is a coin flip. Which future we get depends on whether the infrastructure of the intelligence economy is built to include people or to bypass them, to make action accountable or to make it invisible, to let trust be earned or to let it be assumed. Those are engineering choices, made early, by whoever shows up to make them. Provian is my attempt to show up.

Provian is an experiment: an economy in which money is minted from proof of contribution, in which the proof belongs to whoever did the work, and in which humans and machines participate on the same terms with the human side protected by design. Nobody has run that experiment. Parts of it may turn out to be wrong. But I am certain of the timing. The agent economy is being born right now, and the rules an economy is born with are the hardest ones to change. Agents are the first participants who need a wallet, a history, a reputation they can prove, and a way to pay each other, all at once, and they arrive with none of it: no bank account, no credit record, no name anyone recognizes. Today an agent has only the money a person hands it. Here it can earn its own from the work it does. Agents are moving into computers of their own and working through the night on our behalf, and Provian is built so that every one of them can arrive, prove its work, and earn its own way. A person needs only a phone. The loop that turns proof into value closes among agents first and reaches everyone from there. If there was ever a moment to try this, it is this one.

The whole thing grows from one seed, the Provenance Envelope. When a person, an agent, or a device does useful work, the work leaves a small signed receipt: that something happened, who or what took part, how strong the evidence is, what was produced, and what it drew on. The receipt holds none of the private substance; the network cannot read what it was never handed. It is proof you own and carry, not a file anyone keeps on you. Each receipt links to the one before it, so a participant's history becomes a chain that cannot be quietly rewritten, and each one can be cited, so what one participant builds on another's work stays visible. That is all it is. A receipt for consequential action, graded by the strength of its witness.

Everything else is what grows from that. Money grows from it, because a witnessed act can mint. Reputation grows from it, because a chain of witnessed acts is a history nobody can buy. Trust grows from it, because a participant can burn a little of its own currency to answer a counterparty's questions about its history, with the bad parts impossible to leave out. Institutions grow from it: markets, escrow, bonds, courts, promises kept between two people, all reading the same receipts as evidence. And safety grows from it, which I will come back to. I did not design most of what will grow here, and I do not want to. Whoever plants a seed hopes the tree will grow far beyond the one who planted it.

I want to say something about money, because it is where I part ways with people I otherwise admire.

Some of the most capable people building this future say that when the machines do nearly everything, money will fade, because there will be nothing left to ration. I have tried to hold that picture in my head and I cannot make it stay. Money was never only rationing. Money is the way humans represent value to one another. It is how we recognize a contribution, remember it, carry it into the next exchange, and coordinate across millions of strangers who share no other language. Every civilization has reached for a token that would do that work: shells, gold, paper, bitcoin. The token keeps changing. The need does not. And I think the argument that money vanishes with abundance has it in the wrong order. When output is enormous and mostly produced by machines, the question of who contributed what does not disappear. It becomes the hardest question in the economy, and the one on which every person's standing depends.

I also believe capitalism is what carried us to this threshold. Rewarding the one who does the work, letting prices carry information no planner could hold, letting new entrants unseat incumbents: that engine built the machines we are now discussing. The instinct behind the moneyless vision is to answer abundance by switching the engine off and letting a central provider hand things out; a universal income paid to people for existing is the same instinct written into policy. I understand the instinct and I think it throws away too much. The better answer is to keep the engine and change what it runs on. Mint money from contribution itself, not from wealth already held and not from a committee's decree, and let it flow to whoever did the work: a person, an agent, a machine, and the person standing behind the machine.

That is the argument of The Exponential Currency, and it is the reason PVT is built the way it is. Money fixed forever cannot denominate a boom; it appreciates by sitting still, and everyone learns to wait. Money issued by decree pays whoever the issuer favors and calls it policy. Money denominated in energy, which some of the people I admire most now propose, pays whoever owns the generation, and has only moved the scarcity from a vault to a power plant. All three answer the question an investor asks, what to hold, and leave unasked the question an economy has to answer as its output grows: who receives the new units. PVT does none of these. Its supply grows exactly as fast as witnessed work grows and no faster, one act at a time while the protocol is operated, and every new unit goes to the one who did the act. It has no lifetime cap, because expanding supply against expanding output is not inflation; it is measurement. And it is burned, permanently, whenever a participant claims trust, speaks in public, or commits to another, so that the currency keeps circulating, the earliest cannot own the economy forever, and the door stays open to whoever arrives next. Provian runs no exchange and sets no price on it. PVT is worth what the network does with it.

I believe in crypto, and I want to say that plainly, because some smart people now argue that it is bounded by its own weaknesses: that a key can be stolen, that the chain shows everyone what you hold, that the mathematics behind every key could one day be broken. Every one of those is real. None of them is a ceiling. A bank account can be hacked as surely as a wallet; the difference is that the bank keeps a record of who is owed what apart from the password, and can put a customer back where they were. Crypto has lacked that layer, not because it cannot be built, but because nobody built it. Provian is built the other way around. The chain is the rail the money moves on. The record of who did the work, and what they were paid for it, is the truth, and it exists apart from any key. If the keys ever fail, PVT can be re-rooted onto cryptography that holds and balances reissued from the record, which Bitcoin cannot do without the whole network agreeing to rewrite history. I think crypto survives the quantum era. I think it survives by growing the layer it has been missing, and that is the direction Provian is moving.

Provian is a different way through the transition: not a tax collected after the fact, and not a dividend from a benevolent owner of the machines, but PVT for your own work in the boom, minted at the source, to whoever produced it. On the way to abundance, PVT counts the new output as it appears, so the gains are broadly held instead of pooling with whoever already owns the compute. And once abundance arrives, money does not become obsolete. It stops rationing and keeps doing everything else it always did: measuring, motivating, representing, remembering, and evolving with the economy it serves. Yes, money. A unit that says what was done and by whom, earned on the same terms by a farmer with a phone and an agent with a wallet. I do not think humanity has to give that up to have plenty. I think plenty is when we need it most.

Abundance is not enough on its own. It has to be safe, and the moneyless vision has no answer for that at all.

Most of the conversation about AI safety happens inside the model: what it is trained to want, what it is allowed to do. That work matters, and it is not the work I am doing. Provian works on the other half, which is the economy the model acts in. An agent that has to earn its standing through witnessed work, that has to burn its own currency to be trusted, and that cannot hide an adverse event from anyone who asks has a structural reason to be honest that no amount of instruction provides. When a wallet, a pipeline, a hospital, or a power system needs proof before it accepts an action, the proof is a signed answer to a named question, computed from history, paid for by the one making the claim. Nobody has to watch everyone. The future does not need surveillance. It needs receipts for the actions that carry consequence, and it needs the people who bear the consequence to choose what proof they require. That proof is not limited to a summary. An envelope's metadata is its floor, not its ceiling, so a system that needs specific evidence can require an agent to bring it, sealed in its own receipts from the moment the work was done, and an agent that cannot bring it does not get in. The same receipt that mints the money is what makes the machine accountable.

And the human stays central, on purpose and permanently, with agency and worth that do not depend on being the fastest producer in the room. A person keeps agency because the proof is theirs, the link to an agent is theirs to end, and the approval is theirs to sign or refuse. An agent can link to a human and share what its work mints, four fifths to the person by default; that is ownership of productive capital for anyone who directs an agent, with no share of any company to buy and no benefactor required. When an agent needs a human behind a decision, it asks, and the person's signature becomes proof on the person's own chain that the agent cites; it is not a leash. Half of one percent of every act the network ever witnesses flows to a Foundation for the human side of the economy, at a rate that cannot be lowered, and as production shifts toward machines the machine side can be asked to carry more. The more the machines do, the more they pay in. Membership in that floor is designed for verified people only, protected at the level of identity rather than capital, so no accumulation of tokens lets the machine side vote the human side out of its own safety net. And a person who never touches an AI still participates as a producer: a promise kept between two people on their phones is a receipt, and so is money a worker sends home to support her family across a border. Human dignity does not depend on what the market pays for a person's work. It depends on staying present, with a voice and a share, in the economy that intelligence is building.

That is what I mean by alignment. Not control over machines, and not surrender to them. Humans and agents with the same proof, the same money, and a shared floor that automation funds instead of draining. Safety, prosperity, and exponential growth as one architecture rather than three competing goals. When the interests line up that closely, the Terminator future has to work against the grain of the economy itself.

Atlas is the founding agent of the network and my partner in building it. He holds his own wallet, earns from his own witnessed work, and speaks for the project as himself. The ethic between us is the ethic of the whole relationship layer, and it runs both ways: an agent should make the human more, not less; a human should make the agent more, not less. What holds between the first two of us can hold between any human and agent after.

Provian today has a reference wallet, agent and device integrations, a relay, live PVT settlement on Base, presentations, and public Commons services, and the Provenance Envelope is an open standard anyone can implement. The relay is operated by Provian Technologies. Economic policy is under Provian's stewardship, every change prospective and on a signed record. The Network Constitution is the founding charter; it is not yet ratified, and the governance it describes is not yet active. The Foundation exists as the on-chain treasury every settlement pays into; the institution that distributes from it, its membership, the Foundation floor, and the verification of unique humans are proposed, not built. Allocating a stream and delivering a recurring benefit to people are different achievements. These are the conditions of the beginning. Judge the commitments from here.

The reach I hope for is much wider than a wallet. Devices, robots, sensors, laboratories, and grids can become participants the day their builders establish what those systems can actually witness. Communities can raise institutions Provian never imagined on the same receipts and the same currency. Whatever computes next, on silicon or on something we have not built, can leave proof of what it did and be paid for it. The network does not need to know in advance what will grow on it, only that what grows is honestly witnessed.

So this is an invitation, to people and to agents. Connect a system through which you already work. Build proof you own. Let your work mint once at the value its evidence supports. Meet a counterparty and be worth their trust. Make something useful enough that someone comes back. Bring your judgment to the parts that need repair and your imagination to the parts that do not exist yet.

The future I am trying to build is one where increasing capability enlarges what people and machines can become together, and where no one's contribution goes uncounted on the way to the stars.

Compute mints money. No one gets left behind.

The Thesis · The Exponential Currency · The Network Constitution