# Qubic: The Long Version (QC, Qubic Church)

## Qubic: The Long Version — Chapter 9. The Breathing Token

Source: https://qubic.church/books/qubic-the-long-version/09-the-breathing-token
Author: QC · Publisher: Qubic Church · 2026

### Inhale, exhale, week

Let’s start with a simple reminder. A Qubic epoch is one week. Every Wednesday, around twelve o’clock UTC, the network restarts, 225 of the 676 computor seats can change hands, the counters reset, and a new round begins. We covered this mechanic in Chapter 7; here it’s needed as the backdrop for one specific thing: with that same restart, every week, one trillion fresh QU enters the network.

Not “about a trillion.” Exactly a trillion. This is the network’s inhale. Then, over the seven days until the next Wednesday, the exhale happens. QU are burned—through several independent paths, each with its own purpose, each with its own logic. Some are hardwired into the protocol, some depend on decisions by the computors, some depend on outside economics the network has no direct control over. By the end of the epoch, the balance of inhale and exhale gives the net change in circulating supply. In 2025–2026, this balance regularly came out negative: the network burned more than it issued.

Hence the working name of this chapter. The token that breathes. Not “deflationary,” not “inflationary”—breathing. Every week one inhale; over the course of the week, a series of exhales of varying depth. Supply pulses in time with the epochs. In Chapter 7 I used the word “pulse” for the rhythm of the network. In this chapter the related physiological metaphor returns—breathing. That’s a deliberate choice. In Chapter 8 the register was industrial—gas, pipe, turbine, multi-stage extraction; the chapter was about energy. Chapter 9 is about economics, and Qubic’s economy is built as circulation, not as combustion. So physiology is the more accurate fit than industry. Heart and lungs—two rhythms of the same organism; the network’s weekly rhythm shows up in both.

But I won’t overplay the metaphor. The image should work, not be played with. Now—about the contrast that hangs over this whole topic.

In 2008, the anonymous author of Bitcoin made one decision about supply, and it hasn’t been up for discussion since. Twenty-one million coins, halvings every four years, final issuance somewhere around 2140. This isn’t a matter of voting or a parameter you adjust along the way—it’s a constant baked into the code at network launch. A Bitcoin holder knows: however many years go by, however conditions change, supply will be that. That is the value proposition.

In Qubic, supply is debated every Wednesday. 676 computors vote on how much QU should remain in circulation by the end of this epoch. They can decide to burn more or less. They can—in theory—decide to change emission parameters for the next epoch. Everything having to do with supply lives inside a rolling voting process, not inside the up-front decision of one person.

This isn’t a small technical difference. These are structurally different answers to the question “what is money.” We’ll come to a full statement of the difference in Section 6 of this chapter, when we work through the central quote from the system’s creator. Three questions:

First: where do QU come from? Section on emission.

Second: where do they go? Section on the seven burn paths.

Third: who decides, and how, how many should be in circulation? Section on the computor vote, the link to Axelrod’s fourth condition, and the Arbitrator’s role as a backstop.

Let’s start with the inhale.

### Inhale: a trillion every Wednesday

Each epoch the network issues exactly one trillion fresh QU. This is a protocol constant, hardcoded, not adjustable by vote. It is the network’s baseline “salary,” its fundamental input. In annualized terms—fifty-two trillion QU (52 × 1T = 52T gross issuance per year).

This one trillion is split among 676 computors—those holding active seats in the network for this epoch. A simple division gives an important constant that already came up in the previous chapter: Issuance Per Computor, IPC = 1,479,289,940 QU. One trillion divided by 676, rounded. In Chapter 8 this number figured as a parameter in the revenue formula. In Chapter 9 it picks up its economic meaning: it’s the upper bound on what a computor can earn in a single ideal epoch.

“Upper bound”—not “guaranteed amount.” Real income is computed from the formula in Chapter 8: revenue = IPC × M × (S² + B·E) / DIVISOR Just to recap the structure: the mandatory component M (TX relay, oracles) multiplies the rest of the formula, and the external mining component E gets added in through a small additive bonus. From this followed the empirical observation in Chapter 8: a computor that does only the required network work and UPoW, but doesn’t engage in external mining, gets roughly 80% of IPC. With full external mining—100%. That’s the same 80/20 split we met as a fact in Chapter 8 and that we’ll come back to in Section 3 of this one.

Now—an important point about how distribution works. The system’s creator states the principle of issuance briefly and without euphemisms—he calls it subsidy. In a public message from May 20, 2024:

“Why bother if subsidy method is widely used in everyday life? When you buy a ticket to ride 2km by bus you pay same as one who rides 7km.”

The analogy is clear. On a bus, you pay for the ticket, not for the kilometers. Ride two stops or seven—the price is the same. Many utility schemes work this way, subscription services work this way, subsidized healthcare works this way. The idea of subsidy is a flat price for the fact of participation, not a price proportional to volume of consumption.

In Qubic this is applied to the distribution of issuance. All 676 computors get a comparable share of the trillion—as long as they remain active and above the threshold. There is no strong correlation between how many extra digests a computor found above the quorum threshold and how much they receive. Above threshold—you get your IPC × (coefficients). Below threshold—you risk dropping out in the 225 rotation and losing everything. Inside the active zone—subsidy.

Why this way. It isn’t a gesture of generosity. It’s a mechanism for preventing cartelization. If issuance were distributed proportionally, like Bitcoin proportional hashrate, where more hashrate gives a larger share of the reward—Qubic would instantly switch on the same industrial-scale cycle that turned Bitcoin mining from a million private devices into three or four big pools. Big capital goes into specialized hardware; specialized hardware beats commodity hardware; economies of scale push out the small players; concentration grows. With subsidy, that cycle doesn’t switch on. Scaling hardware to industrial levels in Qubic is pointless—there will be no proportional reward. It’s enough to be a “good” computor to get your share; being the “best” doesn’t grant you a disproportionate bonus.

There’s a direct link here to Chapter 3. Axelrod’s fourth condition: winning the game must not be convertible into disproportionate influence. Translated into the space of economics, this means it must not be convertible into a disproportionate share of issuance. Subsidy performs exactly that function at the distribution level. A fine-print caveat, to keep the picture accurate. Subsidy in Qubic isn’t absolute. The same revenue V2 formula I gave above is not flat. In the multiplicative component M, a computor who handles TX relay badly or sabotages oracles will get less. A computor who actively does external mining gets a +20% bump through the additive term. But these “corridors” are narrow. Inside them—flat subsidy; between them, the difference is not “a hundredfold,” it’s “twenty percent.”

And one last important property of this inhale. No one personally owns the trillion QU that comes into the network every Wednesday. There is no “team allocation,” no “vested tokens,” no founder’s treasury. The trillion gets distributed via IPC according to that week’s activity—made it into 676, did your work well, got your share. Next week— from scratch, no accumulated privileges. Whoever didn’t make it into 676 got nothing. Whoever did got roughly the same as everyone else.

That’s one inhale. But IPC is only the moment of inhale. Across the whole week leading up to the next Wednesday, the exhale runs, and it has several simultaneous directions.

### Exhale: seven ways to burn QU

If issuance is one mechanism, one trillion a week, rigidly defined by the protocol, burning is structured differently. Burn is seven parallel paths, each with its own purpose, its own economic function, and its own degree of predictability. Below—one short item per path.

Path one. Computor vote. Each epoch, the 676 computors vote on burn rate parameters—how much of circulating supply should be burned by the end of the epoch. This is baseline deflation, the constant background exhale. The system’s creator, in a very early publication on January 8, 2022, formulated the logic plainly:

“If computors increase burn rate they’ll make everyone’s QUs to be worth more.” The phrase is simple and free of euphemism. Burning is a redistribution of value from issuance to existing holders. Not an “ecological measure” and not “charity”—an economic mechanism in which computors decide how much new issuance the network leaves in circulation and how much it withdraws. The decision is made every week and every week can be different.

Path two. Smart contract fees. All QU spent by users to execute smart contracts on Qubic are burned. They don’t go to computors, aren’t set aside in a treasury—they go to zero. A quote from October 14, 2021, before mainnet launch:

“All QUs spent to run smart contracts will be burned.”

This is a structural difference from Ethereum. In Ethereum, gas fees (after EIP-1559, with the base fee burned, but the tip—not) partly go to validators; in Qubic, fully into burn.

Path three. IPO bid burns. When a new smart contract comes to the network, it goes through an IPO procedure—computors bid for shares of the future contract. The bids temporarily withdraw QU from circulation. CfB’s formulation from July 5, 2023: “CONTRACT_IPO_BID transactions will be changing total supply (because bid amounts are 'extracted’ from it), in the end of the epoch all bids should be released and restore the initial total supply.”

By the end of the epoch, most of the bids return to the bidders—but some (failed bids below the quorum median) are burned. IPO is a temporary burn with a partial permanent residue.

Path four. Sending to the AAAA… address. An address consisting entirely of A’s is functionally equivalent to a burn. QU sent there will never be retrieved—no one holds the corresponding private key, because the address wasn’t generated from a key. It’s a public, openly available burn mechanism—if a holder wants to voluntarily reduce supply, all they have to do is send QU to AAAA. A small cultural footnote: in 2022, in the first months after launch, such transfers were done as a public gesture demonstrating commitment to the “breathing” model.

Path five. Forfeited computor rewards. If a computor performs poorly, is offline, or doesn’t make quorum—its forfeited QU don’t turn into “ownerless funds.” They go either into burn or into the Admin fund—the decision is made by vote each epoch. CfB spoke about this at the end of May 2024:

“More money to Admin isn’t necessarily good, my past experience with the IOTA Foundation was pretty negative in this aspect (but valuable) and I cannot guarantee that these qus will be spent for something good for the community.”

A short caveat. The mention of the IOTA Foundation here is not the author’s analogy and not speculation; it’s CfB’s own explanation of why he leans toward burn rather than Admin.

Path six. AI operations. When Aigarth executes operations on the network, those operations burn QU. On April 17, 2024, CfB put it in one short phrase: “AI run on Qubic will be burning supply too.”

In April 2026, at the time of writing this chapter, this path is mostly symbolic in volume. Aigarth isn’t yet running at full capacity—Chapter 8 went into detail on why. But architecturally the path is in place. By 2027, if the planned deploy goes through, it could become significant. Full disclosure on Aigarth—in Chapter 15 of the book; here, the fact is that AI operations are designated as a consumer of supply.

Path seven. Buyback through external mining. The most “physical” path—and structurally the most interesting. Starting with Phase 3, computors optionally mine DOGE. The mined DOGE goes through pool operators, gets sold on the market, the proceeds go into buying QU on open exchanges, and the QU bought is burned. The full cycle: compute → external coin → fiat/stable → QU → burn.

The twenty percent of revenue a computor earns for external mining (through the additive B·E term in the revenue V2 formula) is the economic fuel for the buyback. The more computors take part in mining, the more external revenue the pools have, the more QU can be bought back and burned.

Critical honesty. Buyback isn’t guaranteed. It’s an option for computors, not part of the protocol. In Chapter 8, I already cited the empirics: roughly 20 of 676 computors don’t do any external mining, and the network runs fine with that. If tomorrow all 676 simultaneously decided to stop mining—the network would keep functioning as a blockchain, issuance would continue normally, UPoW would keep producing consensus, but buyback would stop. The clean deflation the network showed in 2025–2026 would cease to be. This isn’t a bug and isn’t a vulnerability—it’s design. Buyback is a variable, not a constant.

Moving toward synthesis. Starting in 2025, the Qubic network regularly burns 100% or more of the epoch’s issuance. This is an important empirical fact the system’s creator has commented on more than once. On October 11, 2025:

“Monero is holding pretty well, this epoch we have chance to burn 100% of issued supply.”

And four days later, on October 15, 2025:

“Qubic is designed with 100% of issued supply being sold every epoch.” The first quote is about a specific chance in a specific epoch. The second is about the design of the system as a whole. Together they give a formulation worth remembering: with sufficient external yield through the buyback path, the network is capable of burning the entire week’s issuance plus part of accumulated supply. Net deflation becomes an observable property.

But it’s not a law. It’s a possibility that switches on under certain conditions. When the external market is generous—when the price of DOGE is high, when difficulty is in the right range, when the pools are efficient—the exhale is strong, and the balance goes negative. When the market is stingy—the exhale is weaker, and the balance can come out positive. Which brings the inflation conversation back. On December 13, 2025, CfB said a short phrase that’s easy to quote and easy to misread:

“Annual inflation is 15%.”

And right after, in the same discussion, for comparison:

“5% is world’s annual inflation.”

This needs to be read carefully. 15% is gross nominal inflation. It’s what the network adds to supply per year if there were no burns at all. For comparison—fiat inflation worldwide is about 5%. The 15% figure sounds alarming until you put it next to actual net dynamics: with the burn paths working, the network was regularly in net deflation. But you also shouldn’t say “Qubic is deflationary” as a general characterization—that’s not true. The precise formulation: gross inflation 15% per year, net—variable, depending on the week.

We’ve described the inhale. We’ve described the seven exhale paths. But—who decides how the network breathes in a given week?

### Who decides how it breathes

A simple question on the surface, complex on closer inspection. If the seven burn paths are arranged so differently—some automatic (SC fees, IPO), some voluntary (send to AAAA), some conditioned on outside economics (buyback)—then who ultimately makes the decision about how the network exhales in a given epoch?

The answer: 676 computors.

Every Wednesday, at the epoch transition, the computors vote on burn rate parameters. This isn’t a one-time decision made once and frozen. It’s a rolling decision, retaken every week. Everything to do with supply lives inside this vote: how much to burn out of circulation, how to handle forfeited computor rewards (burn or Admin fund), what the parameters should be for the next epoch.

This formulation is the central idea of the chapter, and in the next section I’ll unfold it through a direct quote from the system’s creator. Here—its consequence, which has implications for the whole architecture.

The supply vote is tied to computor status, not to QU holdings.

This is the payoff of Axelrod’s fourth condition, until now only partly played out: through equal voting weight, and through a rotation that does not convert accumulated holdings into the speed of in-and-out. Here it closes—structurally and finally. A QU holder may own one, a thousand, or a trillion tokens. That gives them no vote on supply parameters. Zero votes. To get a vote, you have to be a computor—and to be a computor, you have to perform UPoW and stay in the top 676 by score. The vote is earned through work, not through capital.

And second: a computor’s vote is one of 676, not proportional to anything. A computor who produced 30% more valid digests than another votes the same—one vote. A computor who owns more QU (if they hold any QU at all beyond a working balance)— votes one vote. Structurally the whole thing reduces to the principle “one computor— one vote.”

A consequence for the ecosystem: those who got richer from price appreciation didn’t get more influential. Ten percent of supply in your pocket—zero votes on tokenomics. This is the opposite of what happens in a typical proof-of-stake blockchain, where increasing stake increases voting weight—and through that creates a direct link between economic gain and the right to change the rules. In Qubic, that link is broken at the architectural level. Not by social contract—by code.

Second consequence: there’s no founder with a treasury. CfB has no privileged supply. There are no team vested tokens that unlock in three years and crash the market. There’s no dev fund of a million QU voted on quarterly. Issuance is 1T per epoch, owned by no one personally as an “allocation”; it’s distributed through IPC according to that week’s activity, and if the network’s data for that week doesn’t hand the trillion to anyone in particular—it burns off in rotation balances. Issuance works per-capita activity, not per-capita history.

Now—the only mention of the Arbitrator in this chapter. It’s brief and forward-looking.

In Qubic’s architecture, there is one element that goes beyond the 676’s vote. The Arbitrator—a figure with a cryptographic signature capable in certain cases of intervening in the network’s operation. I’ll work through this role fully in Chapter 10, which is titled “The Arbitrator—A Temporary Monarch.” Here one fact is enough, directly bearing on tokenomics. On July 13, 2025, CfB wrote:

“Arbitrator would expropriate IDs of those computors who have voted for supply larger than 200t.”

The mechanic: any computor can submit cryptographic proof that supply has exceeded the threshold of 200 trillion—and the computors who voted for the expansion lose their IDs. This isn’t positive control (the Arbitrator can’t set supply) and isn’t a mechanism for managing issuance. It’s protection against one specific pathological outcome: a hyperinflationary vote. If it ever happens that 451 computors vote for supply greater than 200T—the self-policing mechanism kicks in.

Why exactly 200T—a question for Section 5. Why the Arbitrator—for Chapter 10. A third consequence of voting: breathing is not at a stable equilibrium. In one week, the network may burn more than it issued. In another—less. In a third—call an emergency vote and change the parameters. This is not a design bug, it’s a deliberate property.

At the moment of writing this chapter, an active discussion is underway in the Qubic community about a possible radical solution. The proposal, voiced by CfB in a public conversation on March 20, 2026:

“We can, but why half measures? I suggest to set max supply at 10% of current supply, so price jumps 10x. Keep 10% from those who has been holding the longest.” This isn’t a decision, it’s a proposal. The idea is a reverse split, a large-scale burn that would cut supply by an order of magnitude and reward long-term holders. The discussion is open. As I write these lines, no vote has been held, the outcome is unknown. Honestly: I don’t know how this discussion will end. It’s possible the decision will be made before the book reaches the reader. It’s possible the context will have shifted by publication.

To understand quantitatively what all of this means, we need to look at concrete numbers.

### Arithmetic: what “breathing” means in numbers

This section is for the reader who wants to verify with formulas what was described in words above.

Let’s start with how simple Qubic’s accounting for burn actually is. On July 17, 2024, CfB published it in one line:

“Epoch * 1 trillion − supply = burned amount.”

Absolute simplicity. Take the current epoch number, multiply by a trillion, subtract circulating supply—you get the cumulative amount burned across the entire history of the network. No hidden accounts, no off-chain transfers, no complex formulas—arithmetic any user can do with a calculator and access to the explorer.

CfB himself ran this calculation publicly on July 13, 2024, for epoch 104: “Multiply current epoch number by 1 trillion and subtract Circulating Supply: 104,929,098,116,420 $QUBIC.”

Let’s count: 104 × 1T = 104T. Circulating supply at the time was about 103.9T—meaning the burned amount came to a bit under a trillion. By that point, across 104 epochs, the network had issued 104 trillion and burned about one trillion—minimal burns, the overwhelming majority of issuance still in circulation.

Over the next 18 months, the picture changed. Below—a table of supply landmarks by which the “breathing” can be traced from launch to the moment of writing. Three points on this trajectory—2021, 2024, 2025—correspond to three regimes. Until mid-2024, the network barely burned anything; from mid-2024, with the expansion of functionality, burns kicked in; from 2025, with external mining as a buyback path, burns exceeded issuance.

Now—about caps. In supply discussions in Qubic, three different numbers often come up, and it’s important not to mix them up. CfB on April 19, 2024, put it this way: “Cap is 1000T, 100T is expected supply we’ll eventually settle at because of burns.” And from the quote already given in Section 4, the third boundary is known—200T, the Arbitrator trigger threshold. We end up with the following:

1000T—hard cap. The rigid protocol ceiling, embedded in the code. If the network somehow approaches this number—that already means a pathological state. 200T—intervention threshold. The “red line” beyond which self-policing kicks in: computors who voted for expansion past it lose their IDs.

~100T—expected settle. The expected stabilization point of supply with the burn mechanisms working. Not a guarantee, a model expectation from CfB.

Three levels—three different functions. Hard cap—impossible in principle. Threshold —protection against a specific deviation. Expected settle—where the network ought to land under normal operation.

Back to the 15% inflation. I already said in Section 3 that this is gross, not net. Let me add the quantitative grounding. 52 trillion QU per year of gross issuance is what the network adds to supply with hard 1T/week issuance. The 15% figure is the ratio of this gross to the base the network reached by 2025–2026. With burns running, net can be anything—positive in a weak external market, negative in an active one. One last arithmetic detail—and it has bearing on market sensitivity to burns. On July 19, 2023, CfB stated an observation that gets quoted often:

“We are burning supply of short-term holders, supply of long-time ones doesn’t affect price. 15% less qus is 50% less tradable qus.”

A translation and a parsing. Burns, on average, take supply out of the hands of those who actively trade—short-term holders. Long-term holders keep QU in their wallets and don’t take part in trading. Their supply formally exists, but isn’t on the market. When the network burns 15% of total supply—in the tradable pool, that can mean a 50% reduction, because what’s burned is active circulating capital, not “frozen” holds. This is an analytical observation, not a promise. It structurally explains why price sensitivity to burns can be disproportionately large: burns hit the live market, not the sclerotic holdings pool. This isn’t manipulation and isn’t a defensive thesis—it’s a consequence of how supply is distributed between active and passive holders in a young network.

One caveat about the data in this section. Live statistics—market cap, holder distribution, exchange listings—is snapshot information that will have changed by the time the book is read. I deliberately give structural numbers in the chapter (IPC, 1T, 200T/ 1000T/~100T, 15%) rather than live numbers. For live metrics—the Qubic explorer, onchain snapshots as of a specific epoch. Such links will be in the appendix. I’ve laid out the numbers. Now—a short section on the philosophy, without which these numbers stay merely technical.

### Philosophy: Satoshi vs. Quorum

Two worlds.

In 2008, when the anonymous author of Bitcoin writing the whitepaper chose twenty-one million coins as the hard cap, he made one decision for everyone who would later enter the bitcoin economy. This decision is not a parameter regulated by vote; it is not a pull request that can be debated; it is not a policy that can be changed at the next election. It is a law fixed in the source code and in the root copy of the chain. Any attempt to change it means a hard fork creating a new network with a new coin. The actual supply of Bitcoin is one person’s decision, frozen forever. This is not a critique. The fixing of supply is part of what a Bitcoin holder buys when they buy Bitcoin. The price of that decision has been paid: impossibility of adaptation. The benefit is absolute predictability. Every new participant in Bitcoin enters a game whose rules about money supply cannot be changed, even if the changes would be objectively useful.

In 2023—December 25, a short message—the creator of Qubic stated the opposite approach. This quote is the central quote of the chapter. It needs to be read slowly, in pieces:

“I don’t know how it will be, Quorum decides. Satoshi set supply in stone—I picked a more decentralized method. We expect flat line of total supply.”

Four sentences, each carrying weight.

“I don’t know how it will be”—the system’s creator explicitly does not fix the outcome. This is the direct opposite of Satoshi, who fixed it. Not “it’ll be 100T in five years, I promise”—“I don’t know.”

“Quorum decides”—the current consensus decides. Not a document, not a formula, not the project’s author. The collective vote of 676, every week, anew. “Satoshi set supply in stone—I picked a more decentralized method”—this is the direct formulation, not my paraphrase. Two different methods are designated: freezing (“in stone”) and a more decentralized method (rolling consensus). What exactly “more decentralized” means here—I’ll work through below.

“We expect flat line of total supply”—“we expect a flat line of total supply.” Not a promise, not a target—an expectation. If the actual dynamics go differently, then they’ll go differently.

What does “more decentralized method” mean?

CfB has a specific understanding of decentralization, not reducible to the number of nodes or geographic distribution (those matter on their own, but that’s a different dimension). Decentralization in his formulation is about who can change the rules through a legitimate procedure.

In Bitcoin, the rules about supply cannot be changed. Even if 100% of Bitcoin Core developers agree to raise the cap to 22M—that’s a hard fork, and the social-economic response will reject it as not-Bitcoin. Immutability of supply isn’t just a technical property, it’s a social agreement held in place by the expectation of immutability. And that very rigidity is itself a form of centralization: the rules are frozen in one person’s decision, made in the past for all current and future participants.

In Qubic, the rules about supply are tunable each epoch through a legitimate procedure. 676 computors vote; the vote is documented; anyone can verify the outcome. This doesn’t mean supply will change chaotically—the expectation of plateauing is articulated well enough—but it does mean that a mechanism for change exists and belongs to the network, not to its founder.

The difference: in the first case, the decision is made once and frozen; in the second, it’s remade every week.

The counterargument I’m obliged to give, because it’s strong. One of the key value propositions of Bitcoin is precisely the predictability of supply. “I know there will always be 21M of us”—that statement can’t be made about Qubic. Quorum could, in theory, vote otherwise. Does that mean Qubic’s tokenomics is less predictable and, accordingly, less reliable as a store of value?

The answer is—partly yes, and CfB doesn’t argue with that. He’s making a different choice. In Chapter 3 we worked through the fact that Bitcoin can’t fix its own problems —proof-of-work, hashrate concentration, governance deadlock—precisely because it has no mechanism for adaptation. The rigidity of supply isn’t an exception to that pattern, it’s a particular case of the same inflexibility. Qubic accepts the trade-off: losing “guaranteed supply” in exchange for the network’s ability to respond to outside conditions.

Which choice is better—that’s a values question. It isn’t engineering and it isn’t math; it’s a position. The position of this book is this: the trade-off Qubic makes makes its tokenomics alive. Bitcoin answers the question “what is money” like this: money is the impossibility of issuance beyond what was agreed. Qubic answers like this: money is coordination fuel, adjusting to the activity of the network.

Both answers are defensible. Both have economically coherent supporters. This book is about Qubic, and so it describes the second answer. But I’m not saying the first is wrong. It’s different.

A short comparative table, to put the differences into one picture:

Close this section with a short phrase. The question isn’t which system is correct. The question is which answer to the original question about the nature of money is chosen. Bitcoin chose one answer—and built the best possible implementation of that philosophy on it. Qubic chose another—and is building the best possible implementation of this one. They aren’t enemies and aren’t alternatives in an exclusive sense. They’re about different things.

### Breathing closed, Arbitrator unaddressed

To bring the chapter together.

Issuance—one trillion QU per week, hard and undisputed. This is the network’s inhale, its fundamental input. Distribution—subsidy: all 676 computors get a comparable share, inside narrow corridors, with no industrial-scale economies. Exhale—seven burn paths, from automatic (SC fees) to externally market-conditioned (buyback through DOGE). In 2025–2026, the network regularly burned 100% or more of an epoch’s issuance—clean deflation as an observable property when the buyback paths are working. The decision about how the network breathes in a given epoch is made by 676 computors—through a vote refreshed every Wednesday. The vote belongs to the computor, not to the QU holder. This is the structural payoff of Axelrod’s fourth condition, played out at the level of tokenomics: winning the game (accumulating QU) doesn’t convert into the right to change the rules of the game (a vote on supply). The Arbitrator is a separate figure, mentioned here once as a backstop against the pathological 200T+ scenario; full disclosure in the next chapter.

Supply control in Qubic lives in current consensus, not in a past decision. This is the core philosophical choice, set against the Bitcoin model of fixed supply. “Satoshi set supply in stone—I picked a more decentralized method,” the creator of the system stated at the end of 2023. Breathing has no stable equilibrium; it fluctuates with network activity, the external market, the computors’ vote, discussions in the community. April 2026 is the time of one such discussion: about a possible radical reverse split. Its outcome is undecided.

Breathing is a living process. There is no stability in it. This isn’t a bug, it’s a property of the design. The living breathes; the dead has fixed lung volume.

Chapter 6 gave the network’s space: 676 votes, structural form.

Chapter 7 gave its time: weekly epochs, the rhythm of transitions.

Chapter 8 gave its energy: UPoW, the work that produces consensus and accompanying outputs.

Chapter 9 gave its economy: breathing, what circulates between inhale and exhale. Four dimensions described. One left—the fifth. Institution.

In all the architecture I’ve described over the past four chapters—676 votes, weekly epochs, UPoW energy, breathing supply—there is one figure that can intervene. The Arbitrator. In this chapter he appeared twice—once as a backstop against supply >200T, once as a forward reference set out for later. In both cases—in the role of a fuse, not a ruler. But if the network is built as decentralized as I’ve described—why one figure with veto rights at all?

And one more question. If CfB chose a “more decentralized method” for supply, if for years he has consistently handed decisions to the collective and stated this himself as a principle—why has he kept this particular role for himself?

The answer is in the next chapter. The second part of the question, which hasn’t yet been voiced in this chapter, is there too.

The next chapter is about the one whose signature can stop everything. And about why he promised to destroy that signature.