This blog has spent post after post making the case for hard money — for scarcity that no state can dilute, for the architecture that protects a society from the silent tax of inflation. So it would seem perverse to now sketch a money designed to inflate on purpose. The perversity dissolves once you see what the exercise is for. The point is not to propose a rival to Bitcoin. The point is that designing Bitcoin’s deliberate opposite is a precision instrument for understanding Bitcoin — and the first thing the instrument reveals is that the property everyone praises and the property everyone complains about are the same property, seen from two sides.
The Paradox at the Core
Money is asked to do two jobs. It should store value — carry purchasing power forward in time intact. And it should serve as a medium of exchange — move from hand to hand to settle trade. The trouble is that these jobs make opposite demands on the one variable that matters most: how a unit’s value behaves over time. A good store of value should appreciate, or at least hold. A good medium of exchange should circulate freely — which means its holders should have no strong reason to cling to it. And there is the trap: a money that reliably gains value gives every holder a rational reason not to spend it. Why pay today with something worth more tomorrow?
This is not a quirk of Bitcoin; it is a law that Bitcoin happens to express with unusual purity. Its uncompromising scarcity — twenty-one million, the halving driving issuance toward zero — is the engine of its strength as a store of value, and that same scarcity is precisely why it tends to be held rather than spent. The feature you most want to preserve directly produces the behaviour you would need to suppress to make it good cash. You cannot tune one without detuning the other, because they are the same dial. The old name for the visible half of this is Gresham’s law: soft money circulates, hard money vanishes into the savings jar. Bitcoin going into cold storage rather than corner-shop tills is Gresham operating exactly as advertised.
The property you want to keep produces the property you want to lose. They are not two dials. They are one dial, read from opposite ends.
The Instrument: Designing the Opposite
So picture the missing piece — a companion money built to flow. Not a replacement for digital gold, but the spending layer beside the saving layer: hard money to hold, soft money to pay with. The instant you try to actually specify it, every choice teaches you something about why Bitcoin is built the way it is.
Start with issuance. Bitcoin’s halving drives new supply toward zero — the source of its hardness. The soft money would need the opposite: a permanent, gentle dilution that makes hoarding mildly unattractive, so the money keeps moving. Say a constant percentage — on the order of one to two percent — written into the protocol as a formula every node computes deterministically from the current supply. No discretion, no committee, no central bank deciding how much to print. And here the instrument delivers its first sharp reading: this is exactly where the design pays an honest price. Bitcoin’s “twenty-one million” is a single, unarguable number, and its very simplicity is a security feature — it makes any attempt to change it socially impossible. “Constant two percent” is a chosen parameter, defensible but arguable — why two and not one-point-five? The hardness lost there is real, and it is the price of wanting this money to be money and not gold.
What Consensus Actually Requires
The deeper lesson comes from asking what such a money would need to stay trustworthy — and it is the part most people skip when they imagine “just make a faster coin.” The issuance can be a mere formula. Consensus cannot. The reason is the double-spend problem, and it has two layers that are constantly confused.
This is why a soft money built for speed could not quietly drop proof of work to save energy. Strip it out and the question “who chooses the canonical order of transactions?” has only three possible answers: energy (proof of work), capital (proof of stake), or an authority (a central operator). The third is the forbidden central bank by another name. The second imports “those who have, decide,” plus a bootstrapping chicken-and-egg. That leaves energy. The instrument has just demonstrated, from the opposite direction, why Bitcoin’s most criticised feature — its energy use — is not incidental but constitutive. Any money that wants no ruler must pay for its consensus in something no one can fake, and energy is the only candidate that survives scrutiny. It is the same conclusion The Unproven Lock reached from the side of mathematics: the anchor is the incentive architecture, not any single replaceable part.
Where the Real Power — and the Real Threat — Lives
Follow the instrument one step further and it locates the genuine vulnerability, which is not issuance but censorship. If miners receive the entire issuance as block reward, they are recipients of the money supply, not steerers of it — the amount is fixed in the formula, so no miner can print more by being greedy. The coins enter circulation automatically through cost pressure: mining burns real money in electricity and hardware while earning the coin, so miners must continually sell to cover costs. The cost pressure is the distribution rule; no extra mechanism is needed.
The residual power that does remain is the ability to choose which transactions go into a block — and therefore the ability to censor. The realistic threat is not too few miners but too few deciders over block content. Miners cluster into pools to smooth their income, and in the classic model the pool operator picks the transactions — concentrating censorship power into a handful of companies with licences and bank accounts, which is to say, attackable entities. The countermeasures are instructive precisely because they show what a clean design would build in from the start: protocols that hand transaction selection back to the individual miner while letting the pool keep only the income-smoothing, and optional, selective privacy — making transactions hard to single out, so a censor cannot target what it cannot identify. Privacy here is a censorship wall, not a slogan: it works even against a hostile majority that simply does not know what to filter.
The Wall the Instrument Hits
And then the thought experiment runs into something it cannot design around, which is the most honest moment in the whole exercise. A soft money like this would have to bootstrap the way Bitcoin did — starting worthless, mined into existence, gaining value only through voluntary acceptance. But it would bootstrap with a handicap Bitcoin never had. Bitcoin’s early holders had a speculative engine: they held through the worthless years betting on appreciation, and that bet carried it. A money designed to inflate has deliberately switched that engine off — no one holds it speculatively, by design — so it must draw its early value purely from usefulness. Which raises the one question the design cannot answer: what persuades the very first merchant to accept it, before any network exists, on a day when it competes against already-accepted Bitcoin and universally-accepted fiat? That first-acceptor spark is a genuine chicken-and-egg, and the honest move is to mark it as unsolved rather than paper over it. The instrument, in other words, ends by measuring its own limit — which is exactly what a good instrument should do.
What It Means
This was not run to advertise a coin — there is no coin — but because building Bitcoin’s deliberate opposite on paper is one of the most clarifying things you can do to understand the original. Three readings survive the exercise.
Store of value and medium of exchange are one dial, not two. The central finding is structural, not specific to Bitcoin: you cannot maximise both monetary functions in a single property, because reliable appreciation is precisely what discourages spending. Bitcoin expresses the law in its purest form, and understanding it as a law rather than a defect to be patched is the whole point. Energy use, in turn, is constitutive, not incidental — any ruler-free money must buy its consensus with something unfakeable, and of energy, capital, or an authority, only energy avoids reintroducing the very centralisation the money exists to escape. This is exactly the disease diagnosed in The Cathedral Problem seen from the engineering side: hard money is what removes the printing press, and the press cannot be removed without paying, somewhere, for trust.
And the honest gap is the most valuable output. The design hits a wall it cannot clear — the first-acceptor bootstrapping problem — and the discipline is to name it plainly rather than wave it through. That same honesty is what separates a serious monetary argument from a whitepaper selling a token. The exercise is worth more for the wall it found than for the machinery it sketched.
Flight Log — Dispatch from Altitude
There is a number every pilot learns to respect early, and it is a single dial that points two ways at once: airspeed. You would think faster is simply better — more speed, more lift, more margin. It is not. Every wing has a speed that buys the most lift for the least drag, the speed at which the aircraft is most efficient, and on either side of it the picture worsens. Too slow and you approach the stall; too fast and drag climbs and the structure strains. The very thing you want more of in one regime is the thing you must shed in another. One dial, read from opposite ends.
The deeper lesson is that no single airspeed is optimal for everything an aircraft must do. The speed that climbs best is not the speed that ranges farthest, which is not the speed that endures longest aloft, which is not the speed you cross the threshold at on landing. A pilot does not search for one magic number that does all jobs at once — that number does not exist. They learn the speeds, plural, and fly the right one for the task in front of them. Trying to force one speed to serve every phase is how you arrive too fast to land and too slow to climb.
Money is asking the same question that airspeed answers, and getting the same answer. There is no single monetary speed that stores value best and circulates best at once, for the same reason there is no airspeed that climbs best and lands best at once. The properties that optimise one function detune the other, because they are settings on the same dial. The hoarder’s paradox is just the monetary form of a thing every cockpit already knows: a system asked to do two opposed jobs does not want one perfect setting — it wants the honesty to admit there are two settings, and the judgement to use the right one for the phase you are in.
And the part the instrument could not solve — the cold start, the first merchant, the empty network — has its mirror on the ramp too. Every flight begins with an aircraft at rest, full of potential and producing no lift whatsoever, because lift comes only from motion that has not yet begun. The takeoff roll is the riskiest, least graceful phase of any journey: committed, accelerating, not yet flying. Some designs never make it off the ground. The ones that fly are not the ones that solved that problem cleverly — they are the ones that built up enough speed, honestly, one metre of runway at a time. Whether a new money ever reaches that speed is not a thing a protocol can guarantee. It is a thing the runway decides.