The Energy Standard
Money Backed by Physics
There’s a chart at WTFHappenedin1971.com that nobody who sees it forgets. It shows real wages and productivity plotted together from 1947 onward, both indexed so you can compare their movement. For the first twenty-five years they track almost perfectly in lockstep — the economy grows, workers share in that growth, the postwar boom is real and broadly distributed. Then, in 1971, they separate. Productivity keeps climbing. Wages flatten. They haven’t re-converged in the fifty years since. The chart is a kind of X-ray of something that happened to the American economy that no standard explanation quite captures.
What happened is well-known: in August 1971, Richard Nixon went on television and told Americans that the dollar would no longer be exchangeable for gold at the fixed rate of thirty-five dollars per ounce. He called it a temporary measure. It’s still in place.
The standard story is that this caused inflation, which eroded real wages, which produced the divergence. That’s not wrong, but it misdiagnoses the mechanism — and the mechanism is everything if you’re trying to understand what comes next. Around 1730, an Irish-French economist named Richard Cantillon described what actually happens when new money is created: it doesn’t arrive uniformly across the economy. New money enters through specific channels — banks, financial institutions, asset markets — and by the time it has worked its way down to consumer prices and wages, assets have already re-rated upward. People who hold assets gain purchasing power before the inflation arrives. People who hold wages and savings lose it afterward. The same monetary event affects different people in opposite directions, at different magnitudes, at different times.
Balaji Srinivasan, whose thinking on monetary systems has become increasingly useful as conventional frameworks have stopped explaining observed reality, makes the geometric version of this point: inflation isn’t a scalar. It’s a vector. A single policy decision fans outward in different directions for different people — and for different sectors. Plot inflation by category over the past fifty years and you see two economies running inside one number: everything the state subsidizes — healthcare, education, housing — has inflated far beyond general price levels, moving steadily out of reach of anyone without government ties. Everything the market touches — electronics, clothing, manufactured goods — has deflated, driven down by competition, China, and the internet. The “inflation rate” averages these together into a single number that simultaneously conceals both dynamics. What it’s measuring is a directional transfer: purchasing power flows from the people furthest from money creation to those closest, and from the sectors most exposed to competition to those most insulated by subsidy. The 1971 chart isn’t showing that workers became less productive. It’s showing this vector running across five decades, compounding. Money was created. It went, systematically, to the people and sectors closest to the printer. The rules that had constrained this — gold convertibility, the discipline that forced some reckoning with what was being printed — were dissolved in an afternoon on television.
Ray Dalio, who spent his career at Bridgewater studying the rise and fall of reserve currencies across five centuries, has documented what comes next. The sequence runs the same way each time: debt accumulates, debt service crowds out productive investment, growth slows, debasement begins, social fractures deepen, the system resets — through jubilee, collapse, or war. The timing varies. The sequence doesn’t. The Dutch ran it. The British ran it. The United States is now deep in the same pattern. Interest payments are consuming more of the federal budget than national defense for the first time in modern history. The divergence visible in the 1971 chart is the early part of this sequence. The part that comes next is what should concern us — because this time, unlike Argentina or Weimar or the British pound’s exit from reserve status after World War II, there’s no floor. No other monetary system is waiting to absorb the failure. If the dollar loses reserve status without a successor architecture in place, there’s nothing to fall back to.
The obvious candidates to replace it aren’t. The euro is a monetary union without a fiscal union — it doesn’t function as a true reserve currency without collective debt issuance, which Germany has consistently blocked. China’s renminbi isn’t freely convertible; you can’t hold it as a global reserve without trusting that Beijing will let you repatriate it, and that trust doesn’t exist at scale. China’s own debt position — shadow banking, local government financing vehicles, the property sector collapse — is arguably worse than America’s, on a faster clock, with less institutional credibility. BRICS nations are building settlement mechanisms to reduce dollar dependence — but shifting control from Washington to a collective of sovereigns doesn’t change the extraction logic. It changes the beneficiary. That’s the problem. The problem isn’t which sovereign controls the reserve currency. The problem is that any sovereign controlling it will eventually run the same extraction logic, because the substrate rewards it. What’s needed is a non-sovereign monetary substrate — and nothing of that kind has ever existed at global scale.
Understanding what such a thing could look like requires going further back than 1971, to the deeper history of what money actually is, and what it has always been anchored to.
What Money Has Always Been Anchored To
David Graeber, the anthropologist who spent a decade writing *Debt: The First 5,000 Years*, dismantled a story that nearly every economics education begins with. The standard account, going back to Adam Smith, is a simple progression: first humans bartered goods directly, then they invented commodity money to make barter more efficient, then they developed credit instruments on top of that foundation. Graeber’s finding — drawing on archaeologists, historians, and anthropologists across dozens of cultures — was that the sequence is almost exactly reversed. Mesopotamian temple credit tabs from around 3500 BCE predate coins by three thousand years. Medieval Europe ran sophisticated credit systems for centuries during periods when physical coin was genuinely scarce. Coins appear historically not from the organic evolution of markets but from armies — coinage is correlated with state violence and the need to pay soldiers quickly and in portable, standardized form across large territories.
What this means is that money has never been “backed by nothing.” The question isn’t whether it has an anchor — it always does. The question is what the anchor is, who controls it, and whether the verification technology to enforce that backing actually exists.
That last part matters more than it sounds. James Burke, the documentary maker whose career was devoted to tracing how one discovery made another possible, makes the enabling condition for monetary systems precise: commodity coinage only became viable after the discovery of the touchstone — a naturally occurring piece of black siliceous stone, not manufactured but found, whose value was in being recognized. It let you verify a metal’s purity by the color of the streak it left on the surface. Archimedes made the same category of discovery when he noticed that objects displace water in proportion to their volume — a physical fact present in every bath and bucket, waiting for someone to recognize it as a verification instrument. The insight let him determine whether a king’s crown was pure gold or secretly alloyed with silver, without melting it and destroying the object. Both discoveries are instances of the same phenomenon: physical properties of the world that enable trust between strangers once someone notices what they can do.
The inverse case is almost as instructive. The Aztecs and the Inca had gold in abundance — they worked it into extraordinary ceremonial objects, they covered temples with it, they valued it enormously. What they didn’t have was gold as money. In Mesoamerica, the currency was cacao beans, cotton cloth, copper axes. In the Inca Empire, there wasn’t currency in the Western sense at all — an enormous redistributive state ran on labor obligations. Gold was precious. Gold was not money. Without the verification infrastructure — the touchstone, the density measurement, the ability to assay an unknown metal presented by a stranger — gold couldn’t serve as a medium of exchange across people who didn’t know and trust each other personally. The conquistadors who arrived expecting to find monetary gold found instead a civilization that had solved the problem of large-scale coordination without it. The gold they plundered was then processed through European assay technology and turned into the Spanish silver dollar, which then financed empire and eventually triggered the Ming Dynasty’s collapse. Same material. Completely different function. The difference was the verification infrastructure.
Burke’s point generalizes: every monetary leap in history has required a new verification technology — discovered or developed — to make trust between strangers reliable enough for exchange. When that technology didn’t exist, or when it could be circumvented, the monetary system eventually broke down.
Graeber documents the failure mode that commodity money keeps hitting: economies eventually outgrow their money supply. Commerce seizes — not because productive capacity declined, but because the medium of exchange couldn’t scale to match it. England in 1696 provides the starkest exhibit. The silver coins in circulation had been badly degraded — worn thin and clipped at the edges by users skimming small quantities of silver from each one. Two solutions were proposed to a parliament in crisis. William Lowndes, the Treasury official, argued for devaluing the new coins to reflect what silver was actually worth on the market — a pragmatic acknowledgment of reality. John Locke, the philosopher who had recently articulated the natural rights that made governments legitimate, argued for restoring coins to their original silver content at face value: a point of principle, a defense of honest money against debasement. Locke won. Parliament restored the silver standard. The result was deflationary collapse: the money supply contracted violently to match the commodity anchor, hundreds of thousands went out of work, there were riots and widespread hunger, and the economy seized for years. By trying to keep money “honest” in the commodity sense, they made it catastrophic.
The Ming Dynasty ran into the same problem through a different mechanism, and the scale makes it harder to dismiss as a footnote. Having adopted silver as a monetary standard and commuted nearly all taxes to silver payments, the empire became dependent on silver it couldn’t produce domestically. Spanish silver from the mines at Potosí and in Mexico flowed in extraordinary quantities across the Pacific to China to settle those obligations — and when those flows slowed, so did the Chinese economy. Deflationary pressure, social instability, and eventually imperial collapse followed a chain of causation that began with a commodity-backed monetary system hitting the limits of the commodity supply.
Fiat currency — money whose supply is controlled by a central authority rather than tied to a commodity — was supposed to solve this. Make supply elastic. Let money grow with the economy rather than with geological luck and distant silver mines. The diagnosis was correct. The cure introduced a different disease: the Cantillon Effect, now with no commodity constraint to limit how aggressively money could be created or how directly it would flow to those closest to the printer.
The Capture
Before getting to fiat’s contemporary failure mode, there’s an older fraud worth naming precisely, because it predates fiat and sits at the foundation of the modern financial system.
The mechanism begins with a practical problem: physical gold is heavy, dangerous to transport, and easy to steal. A merchant carrying gold across a city — let alone across a country — was a target. By the seventeenth century, goldsmiths had the natural infrastructure to solve this: thick walls, strong locks, guards. People began depositing gold with goldsmiths for safekeeping, in exchange for a paper receipt — a note certifying that the bearer was entitled to withdraw a specific amount of gold on demand. These receipts were backed one-for-one. They were honest instruments, and they were convenient enough that people began trading the receipts themselves rather than withdrawing the gold. Why carry metal when the paper worked just as well and was far easier to transport? The receipts became, in practice, money.
The goldsmiths noticed something. Gold sat in their vaults almost untouched. Deposits came in. Withdrawals went out. But the total in the vault barely moved — people were circulating the paper instead of redeeming the metal. A goldsmith who kept, say, ten percent of deposits as a reserve to cover typical withdrawal demand could lend out the rest, collecting interest. He was issuing more receipts than he had gold. If everyone showed up at once, he couldn’t honor them. But everyone never showed up at once. So the gap was invisible most of the time, and the interest on the conjured portion was real.
The bank that issued ten claims against five units of gold was earning interest on five units it had effectively created from nothing. If anyone else did this it would be called fraud. We gave it a Latin name — fractional reserve banking — a regulatory framework, and called it financial sophistication. Niall Ferguson, the economic historian, makes the honest counterargument in The Ascent of Money: this system really did fund things that mattered. The British Empire’s comparative advantage was partly its ability to issue credible long-term bonds at lower interest rates than its rivals. Railroads, industrialization, the physical infrastructure of the modern world — much of it was debt-financed, and Ferguson is right that credit has driven real productive expansion. But Ferguson also documents where the system goes when nothing constrains it. The Bank for International Settlements estimated the global derivatives market at $846 trillion in June 2025 — roughly eight times world GDP. That ratio isn’t primarily about managing economic risk. It’s a portrait of what invented money looks like when it has nowhere useful to go: real estate bubbles, art markets, collateralized debt obligations, the entire financialized shadow economy that decoupled from productive activity after the 1970s. Ferguson’s bonds funded railroads. The derivatives market funds itself.
In 2007, the BBC aired a documentary called *The Trap*, by the filmmaker Adam Curtis. His subject — which sounds abstract until you understand the mechanism — is how the institutions built specifically to prevent financial extraction were redesigned, from the inside, to facilitate it. But his argument begins much earlier than most people expect.
The post-World War II economic consensus was built explicitly on a reckoning with the Gilded Age. The Depression had demonstrated what unchecked capital accumulation at systemic scale produces: monopoly, mass poverty, spectacular inequality, social fracture, and eventually the political conditions for global war. As Curtis documents, the architects of the postwar settlement said so explicitly — “no longer did we worship at the shrine of no-holds-barred capitalism.” Friedrich Hayek, who believed all of this regulation was worse than the disease, was a fringe figure. The lesson of the first half of the twentieth century seemed clear enough that challenging it required exotic mathematical cover.
That cover arrived in the form of John Nash’s game theory, developed for the RAND Corporation’s nuclear strategists. Nash’s models assumed universal self-interest: that every actor, in every situation, would defect given the opportunity, and that cooperation was always a strategic facade. Curtis traces what happened when this model of human behavior migrated from nuclear deterrence into economics, then into public policy, then into the management of every major institution in the Anglosphere. The lesson of the Gilded Age was gradually forgotten — or rather, systematically discredited by the people who benefited from its forgetting, using the mathematical prestige of game theory as their warrant.
The mechanism, once running, is self-fulfilling. Design institutions around the assumption that actors will defect, and defection becomes the only rational strategy — the model predicts it, the institutional design makes it inevitable, the data confirms it, the model appears validated. Apply the same logic to currency: design money so that those closest to the printer extract purchasing power from everyone else, give them the legal apparatus to do it, and they extract. Not as conspiracy. As Nash equilibrium.
The most consequential instance of this logic wasn’t in a military briefing room or a hospital corridor. It was New York City, 1975.
The city was technically bankrupt. Banks held its short-term debt and refused to roll it over — not primarily because they doubted eventual repayment, but because the condition for refinancing was control. Felix Rohatyn and a consortium of investment banks created the Municipal Assistance Corporation, which took effective authority over New York’s finances away from elected officials and handed it to a board answering to the bond markets. Politicians who had been elected on platforms of public services and labor protections signed away the power to implement those platforms as the price of avoiding immediate collapse.
Adam Curtis, in *HyperNormalisation* — his 2016 follow-up to *The Trap* — argues this moment marked something more significant than a fiscal crisis. Politicians had discovered that when they tried to manage the actual complexity of a modern industrial economy, they failed publicly and visibly. The bankers offered a different arrangement: hand over economic governance and they would manage the complexity that elected officials couldn’t. What was surrendered wasn’t just the city’s finances. It was the operating principle of the postwar settlement — that democratic politics could constrain capital, rather than capital setting the limits within which politics operated.
The growth dimension is what Balaji adds. The postwar settlement worked, for thirty years, because there was a pie growing fast enough that coordination beat extraction. When growth stalled in the 1970s — stagflation, oil shocks, the limits of mass-production manufacturing hitting their floor — the game changed. A stagnating pie converts coordination into extraction by making it zero-sum: what you gain, someone else doesn’t. The generation that built the settlement understood this, because they remembered the Depression. The generation that grew up inside the Golden Age inherited the institutions without understanding their purpose. The guardrails felt like pointless restrictions. The same pattern repeated at the end of the 1990s when the dot-com bust forced a choice between managing complexity or financializing around it. The decision, again, was to financialize.
Balaji’s argument: when you can’t innovate — when you don’t know how to open new frontiers of economic activity — you extract from the frontiers that exist. The innovator doesn’t need the Cantillon machine. If you can build something that didn’t exist before, you capture that value directly. The Cantillon machine is the rent-seeker’s instrument: proximity to money creation substitutes for value production. When growth stalled and the generation that understood the settlement’s purpose gave way to one that had forgotten it, the path of least resistance was to hand economic governance to the people best positioned to extract from existing structures. The Nash defection model spread through institutions not just because it was mathematically elegant but because it gave institutional cover to people who were already planning to defect, in a system that had stopped believing its function was to prevent them.
By the time the 2008 financial crisis arrived, the response to a crisis caused by financial extraction was to hand money directly to the extractors. The distance between the money printer and the wage earner had never been wider. The Gilded Age lesson, carefully institutionalized after the Depression, had been fully forgotten. We got the Gilded Age outcomes again.
The petrodollar was the mechanism that bought time before the reckoning. After 1971, the dollar needed a new anchor. The improvised solution — negotiated between Henry Kissinger and the Saudi royal family — was elegant in its ruthlessness: Saudi Arabia would price oil exclusively in dollars; America would provide security guarantees and military hardware. If the world needs dollars to buy oil, the world holds dollars. That sustained demand absorbed decades of money printing that would otherwise have shown up as faster domestic inflation. Petrodollar surpluses recycled into US Treasuries, funding American debt at favorable rates. This was not a solution to the fiat problem — it was a pressure valve, a mechanism for exporting the consequences of monetary expansion to the rest of the world rather than eliminating them. Billions of dollar-holders globally absorbed American inflation as an involuntary tax.
The valve is failing. The freezing of Russia’s $300 billion in reserves in 2022 sent a message to every sovereign wealth fund on the planet: dollar-denominated assets carry a political risk that wasn’t previously priced in. The neutrality of reserves — the entire premise of the dollar as global reserve currency, the reason people held it rather than something else — was revealed as contingent. BRICS settlement mechanisms, accelerating gold accumulation by central banks, and Chinese yuan cross-border systems are all responses to the same signal. But they are not solutions to the underlying problem. Swapping American sovereignty over the reserve currency for collective BRICS sovereignty doesn’t fix what’s broken — it just changes who benefits from the extraction. The dollar remains the least-bad sovereign option. The point is that no sovereign option is actually good, because the problem is the sovereign.
The Closest Attempt
Bitcoin understood the Cantillon problem precisely. Its pseudonymous creator Satoshi Nakamoto designed it with a fixed supply cap, no central issuer, and a halving schedule — reducing the rate of new coin creation every four years — encoded permanently in the protocol. You cannot debase it. You cannot vote yourself more of it. And Nakamoto solved something subtler than most people notice: gold’s scarcity is geologically contingent. It’s an accident of Earth’s planetary formation. New extraction techniques, or eventually asteroid mining, could disrupt it at civilizational timescales. Bitcoin’s scarcity is enforced by the SHA-256 hash function and the consensus rules of a decentralized network. It doesn’t care what’s in the asteroid belt.
But fixed supply is too deflationary to work as a transaction currency. If your money reliably appreciates, the rational move is to hold it. Every bitcoin spent today is one that would have been worth more tomorrow. This is the Ming trap again, in digital form: the economy outgrows its medium of exchange, commerce strangles, the store of value becomes a drag on productive activity. Bitcoin also already ties money creation to energy — that’s what proof-of-work is — but then burns that energy to prove the point rather than using it productively. The first genuinely energy-backed currency in history wastes the energy as its demonstration.
There’s a deeper problem with Bitcoin that has only recently become documentable, and it matters because the same forces that captured fiat are now capturing crypto.
Satoshi’s original vision — spelled out in the white paper and in his documented communications — was peer-to-peer electronic cash: fast, cheap, borderless, permissionless, usable for everyday transactions outside the control of Visa or Mastercard. From roughly 2012 to 2016, this is what Bitcoin was. Major retailers including Overstock and Expedia accepted it directly. Transactions settled in minutes and cost fractions of a cent.
Then the network was throttled.
Bitcoin has a block size limit — a cap on how many transactions can be processed per ten-minute block. Satoshi added this limit quietly as an early spam-prevention measure, with later communications suggesting it could be raised as the network matured — though the intent was contested enough to fuel years of civil war. For years, the limit was never binding. Then, as adoption grew and the limit started to bite, a faction of developers argued it should not be raised. Bitcoin should be “digital gold” — a store of value — not a transaction medium. Users wanting cheap, fast payments should use second-layer solutions built on top of the constrained base layer. This position happened to be very convenient for a company called Blockstream, which had been founded by several of Bitcoin’s core developers and whose business model depended on building exactly those second-layer solutions.
The block size wars of 2015 to 2017 split the Bitcoin community and ultimately confirmed the small-block position. Transaction fees spiked to twenty dollars or more. Settlement times stretched to days. The payment use case collapsed. Bitcoin Cash — the faction that followed Satoshi’s scaling design — forked off in 2017. Bitcoin Core, the small-block camp, won the narrative war. Ordinary users who discovered that sending bitcoin now cost twenty dollars and took a week had no way of knowing this was a recent, contested, and structurally motivated change to a system that had worked differently before.
What congressional disclosures and DOJ documents released through the Epstein proceedings have made documentable is that this capture wasn’t entirely organic. When the Bitcoin Foundation, the nonprofit that had been funding core developers, collapsed in 2015, MIT’s Media Lab stepped in to absorb them. Emails from Joy Ito, who ran the Media Lab, to Jeffrey Epstein describe the Bitcoin Foundation’s collapse in urgent terms and his plans to find the developers a home. Twenty days after describing the crisis, he confirmed it was resolved: the three developers had joined the Media Lab. “Used gift funds to underwrite this,” he wrote to Epstein, “which allowed us to move quickly and win this round.” Epstein had separately co-invested roughly half a million dollars in Blockstream’s 2014 seed round — approximately a year before the developer funding transition, though the stake was divested within months of the investment — and had endorsed the “digital gold, not digital cash” narrative in a 2017 press interview. Cory Fields, one of the three developers whose transition Ito funded, subsequently worked on Project Hamilton, the MIT-Federal Reserve CBDC pilot, which can process 1.7 million transactions per second, compared to Bitcoin’s seven.
Meanwhile, Tether — the stablecoin whose co-founder Brock Pierce had maintained extensive documented contact with Epstein since 2011, while simultaneously serving as chair of the Bitcoin Foundation at the time it collapsed — was found by the CFTC to have held sufficient reserves on only 27.6% of days during a 26-month sample period — meaning for nearly three-quarters of that time, the claimed dollar-for-dollar backing did not exist. A study by Griffin and Shams found that approximately half of Bitcoin’s 2017 price appreciation was attributable to Tether issuance. The “store of value” narrative required price appreciation to seem credible. Tether — unaudited, fractionally reserved — manufactured that appreciation. It is fractional reserve banking applied to crypto: issue more claims than you have assets, earn the spread, and call it a new financial paradigm.
A system designed to circumvent financial control was first throttled out of its most threatening use case — cheap, permissionless cash — then subjected to the same fractional reserve dynamics it was designed to escape. It’s now providing the regulatory surface area for the Genius Act, which puts stablecoins under financial surveillance, and the Clarity Act, which would put tokenized stocks, bonds, and eventually real property under programmable controls. Whether that’s the intended end state or an opportunistic convergence of interests, the practical result is the same: the most promising attempt to exit the Cantillon machine is being absorbed back into it. And the destination that trajectory points toward is worth examining directly.
The Destination
To understand where the digital money story is heading, you need to understand what makes a Central Bank Digital Currency different from the digital money you already use. When you have a balance in a bank account today, your bank owes you dollars and holds deposits at the Federal Reserve. A CBDC eliminates the bank in that chain. The digital dollar would be a direct liability of the Federal Reserve itself, held in a wallet on your phone, moving across infrastructure the central bank operates and can observe in real time. Every transaction. Every balance. Visible at the source.
That’s different from existing financial surveillance in degree, but degree is not the most important difference. The most important difference is programmability.
A digital dollar issued by the Federal Reserve can be coded with conditions that physical cash categorically cannot be given. Cash is fungible and inert: a twenty-dollar bill is a twenty-dollar bill regardless of who’s spending it, what they’re buying, or whether any authority approves of the purchase. A programmable CBDC is none of those things. It can be given an expiry date — spend it in thirty days or lose it, useful for stimulus the government wants to ensure is actually spent rather than saved. It can be restricted to approved purchase categories — certain food items, domestic goods, things a regulator has designated permissible. It can be made geographically conditional — valid in some regions, worthless in others. And it can be made contingent on a compliance score: the purchasing power of someone who has violated a policy condition can be restricted, reduced, or simply frozen, without the due process that comes with seizing physical property.
China’s digital yuan — the e-CNY — has already implemented versions of this. Regional stimulus distributions have included expiry dates, ensuring the spending multiplier functions as intended. The e-CNY operates over infrastructure that gives the People’s Bank of China transaction-level visibility that physical yuan does not provide. And China’s Social Credit System — which aggregates behavioral data from government and commercial sources to assign scores that determine access to travel, loans, luxury goods, and in some cases educational opportunities for the children of blacklisted individuals — provides the conceptual architecture for connecting financial access to behavioral compliance. The infrastructure for connecting your score to your wallet already exists. The decision to connect them is political, not technical.
Western democracies are further back on this curve, but the direction is visible. Project Hamilton — the MIT-Federal Reserve CBDC pilot whose technical team included Cory Fields, one of the Bitcoin Core developers whose MIT transition Ito funded with Epstein’s gift funds — demonstrated 1.7 million transactions per second. The developer who helped maintain Bitcoin’s seven-transaction-per-second cap helped build the Federal Reserve’s infrastructure that processes two hundred thousand times more. The Genius Act creates the transactional surveillance layer over stablecoins. The Clarity Act extends control architecture to tokenized real-world assets. The legislative scaffolding is going up.
The end state is something that doesn’t have a clean historical precedent. The Cantillon Effect describes extraction through proximity to the money printer — some people receive new money before prices adjust, while the rest absorb the inflation. A programmable CBDC doesn’t just enable that extraction. It enables conditional access. The difference between “the government can freeze your account” — which existing financial infrastructure already permits, through banks and courts — and “the government can determine what you’re allowed to purchase with your own money” is a qualitative leap. The first is a blunt instrument of last resort. The second is behavioral conditioning built into the transaction layer itself. Cash is fungible. Programmable money is a leash with purchasing power attached.
This is what the Bitcoin hijacking narrative points toward. Not a world where peer-to-peer digital cash circulates outside sovereign control — Satoshi’s original vision — but a world where digital money is sovereign money, sovereign money is surveilled, surveilled money is programmable, and programmable money is behavioral management with the infrastructure of convenience. The Gilded Age gave you extraction through labor markets. What’s being built now gives you extraction plus surveillance plus conditional access, through the medium of exchange itself. The Cantillon Effect was the monetary instance of The Trap. The CBDC is the next iteration: defection baked into the substrate, at the level of the transaction.
The Fifty-Year Delay
While monetary systems were failing, a physical constraint was being quietly locked in place.
In 1965, a reactor at Oak Ridge National Laboratory in Tennessee started running. The Molten Salt Reactor Experiment was unlike anything built before. It operated at atmospheric pressure rather than the dangerous high-pressure environments that make conventional reactors so complex and expensive to build safely. The key difference was liquid fuel: conventional solid-fuel reactors swap out their rods at less than one percent utilization because fission products poison the reaction even though most of the fissile material remains. In a molten salt reactor, the fuel is dissolved in liquid — fission products can be removed continuously and the fuel keeps burning. Greater than ninety percent utilization. The design was built to run on thorium, an element so abundant it’s practically a waste product of rare-earth mining, which China is now demonstrating at Wuwei. Its safety model was passive: the reactor’s physics shut it down automatically if something went wrong, through what’s called a negative temperature coefficient — the reaction dampens as it heats up, requiring no active intervention to prevent runaway. Over four years of operation, the MSRE logged more than thirteen thousand hours at full power.
Alvin Weinberg, the director of Oak Ridge who had championed the design, was fired in 1973. The Atomic Energy Commission had committed to a different technology — the Liquid Metal Fast Breeder Reactor, which maintained compatibility with existing uranium and plutonium infrastructure and had the political advantage of remaining entangled with weapons programs. The decision against the Molten Salt Reactor was not technical. The MSRE worked. The knowledge was shelved.
Michael Shellenberger, in *Apocalypse Never*, documents something that has never been fully resolved: in 1969, Robert O. Anderson, CEO of Atlantic Richfield Company, provided funding to David Brower to found Friends of the Earth, which became one of the leading anti-nuclear organizations in the United States. The precise amount is disputed — Shellenberger cites two hundred thousand dollars; other accounts put the figure lower, or suggest the money went to the John Muir Institute rather than directly to Friends of the Earth; Brower denied it. The factual record is murky. The ARCO foundation subsequently funded anti-nuclear campaigns across the US and Europe. The interpretation is contested — oil and electricity were not direct competitors in 1969, and Anderson may have been a genuine environmentalist. The incentive logic is coherent regardless of motivation: solar and wind power are intermittent, and intermittency requires dispatchable backup, and the dispatchable backup that wins commercially is natural gas. Nuclear doesn’t need backup. If you’re in the natural gas business, nuclear competes with you in a way renewables fundamentally don’t.
The pattern extends well beyond a single donor. Fossil fuel companies sponsored pro-solar and pro-wind campaigns throughout the decades when nuclear was being legislated out of existence — not out of environmental conviction, but because intermittent renewables are complementary to their product while nuclear is not. Anti-nuclear advertising systematically fused civilian reactor imagery with weapons iconography: mushroom clouds, Hiroshima, the bomb. A power reactor and a nuclear weapon share fission physics and nothing else — different fuel, different engineering, categorically different risk profile. The public conflation of the two required sustained effort to create. It was provided.
Vernor Vinge, the mathematician and science fiction writer, plotted maximum power source per capita over fifty thousand years on a log-log scale and noticed something most people, looking at the data, don’t quite absorb. The transitions between energy regimes aren’t improvements in degree. They’re phase changes. From early humans at a third of a kilowatt to the steam engine at a megawatt to contemporary power plants at thirteen gigawatts — each step doesn’t just give you more energy. It opens entirely new categories of civilizational capability that didn’t exist below the threshold. We’ve been at roughly the same point on that curve since the 1970s. The knowledge to take the next step existed. The decision was reversed.
Well, that was an unexpected surprise
I did not have nuclear finally making a comeback because of AI power demands on my 2025 bingo card.
AI data centers require 24/7 baseload power at a scale that solar and wind cannot reliably provide. The grid math doesn’t work: you can’t run continuous inference loads on intermittent sources without massive storage infrastructure that doesn’t exist at anything close to the required scale. The result: Microsoft restarted Three Mile Island specifically to power its data centers. Google signed a power purchase agreement with Kairos Power for small modular reactors. Sam Altman — who chairs Oklo, an advanced fission startup — is simultaneously the person most responsible for the extraordinary surge in AI energy demand and an investor in one of the nuclear supply responses to it.
Fifty years of public argument, environmental litigation, and regulatory engagement couldn’t overcome the institutional capture around nuclear power. Academic journals, environmental organizations, regulatory agencies — none of it was sufficient to reverse a decision made in the 1970s by people who had reasons to reverse it. It took a different industry needing baseload power badly enough that the politics became irrelevant. The governance system failed. The economics did an end-run.
China, for its part, ran the MSRE’s successor — the TMSR-LF1 at Wuwei — and in October 2024 loaded thorium and detected protactinium-233, the first sign of nuclear breeding initiating in the thorium fuel cycle. The technology Weinberg was fired for championing is now operational in Gansu province. We spent fifty years not taking the next step on Vinge’s curve. Someone else is taking it.
What Cheap Energy Actually Unlocks
When William Stanley Jevons studied the British coal industry in 1865, he noticed something counterintuitive: as steam engines became more efficient and coal use per unit of output dropped, coal consumption went up, not down. Cheaper energy doesn’t reduce demand — it expands it, by making viable everything that was previously above the cost threshold. Each new application creates its own demand for more energy, which drives further innovation, which makes energy cheaper, which enables more applications. It compounds.
So what does energy at one-tenth today’s cost actually unlock?
Advanced nuclear and geothermal drilling are the two technologies most likely to get there. Nuclear provides the baseload density that solar and wind cannot. Advanced geothermal — applying directed energy drilling to reach depths previously accessible only in volcanic regions — extends the same principle to the heat sitting under the Earth’s crust everywhere on the planet. Two always-on sources, neither dependent on weather or geography, both approaching cost curves that make one-tenth plausible rather than theoretical.
Vinge’s phase-change observation is the frame for what this opens up. It’s not that ten-times cheaper energy makes existing things ten percent better. The cost threshold above which things don’t exist moves, and what’s below that threshold is a qualitatively different world. Specifically:
The chemistry for pulling carbon dioxide from the air and turning it into synthetic hydrocarbon fuel already exists. You capture CO₂, combine it with electrolytic hydrogen, and you get synthetic kerosene, synthetic methane, synthetic rocket propellant — carbon-neutral, manufacturable anywhere there’s energy and atmosphere. The barrier right now is purely economic: it costs more in energy to produce the fuel than the fuel is worth. At one-tenth the energy cost, that relationship inverts. Commercial aircraft run on fuel synthesized from the air. The carbon problem stops being a geopolitical problem — who can afford to cut emissions while growing their economy — and becomes an engineering problem with an economic solution. Not because anyone passed a law, but because the economics changed.
Freshwater scarcity is an energy problem. Reverse osmosis desalination runs at three to four kilowatt-hours per cubic meter. Make that cheap enough and you’re drinking ocean water everywhere, pumping it to deserts. Japan already runs the world’s largest indoor vertical farms — food grown under LED lights, completely independent of weather, season, or soil quality. Scale that with cheap nuclear baseload and the food and water crises become downstream engineering problems, not permanent features of the human condition.
Materials collapse in price when energy does. Aluminum smelting takes fifteen kilowatt-hours per kilogram — aluminum is almost pure stored energy in solid form. When energy becomes cheap enough, structural materials follow, and the cost of everything made of physical things — buildings, vehicles, spacecraft — moves with them. The space fountain, an orbital ring requiring enormous sustained power to maintain against gravity, moves from theoretical exercise to engineering problem. And once manufacturing is viable in space, the accessible materials base changes entirely.
The constraint on civilization right now is not imagination. It never has been. It’s the energy bill.
Why the Climate Crisis Probably Isn’t
Climate change is real. The “unsolvable civilizational crisis” framing is the error — and it matters, because the error is doing serious damage to the generation that has internalized it.
The doomer framing rests on an assumption that almost nobody states explicitly: that the current energy cost structure is permanent, and therefore that the only path runs through political coordination — 195 countries simultaneously accepting economic sacrifice in a game where defection is rational. That assumption is wrong. The engineering path exists, and cheap energy makes it economically inevitable rather than politically aspirational.
At one-tenth today’s energy cost, direct air capture of atmospheric CO₂ — currently $300-1000 per ton, too expensive to deploy at scale — drops to the range where it competes with the damage cost of emissions without subsidy. Synthetic fuel synthesis, already chemically understood, flips from energy-negative to profitable: fossil fuel gets displaced by market economics, not by treaty. Excess nuclear capacity running carbon fixation continuously produces liquid synthetic hydrocarbons as a permanent storage medium for electrical energy — closing the carbon loop as a side effect of doing useful chemistry. The political coordination problem dissolves because the economics solve it first. This is exactly what happened with the Malthusian food crisis: Haber-Bosch nitrogen fixation made the population catastrophe irrelevant not through birth control mandates but through a step-change in agricultural productivity. Nobody negotiated their way out of Malthus. They out-engineered it.
The more serious and genuinely underattended civilizational risk is habitat loss. Species extinction is running at 100 to 1000 times the background rate. The primary driver isn’t temperature — it’s land conversion: agriculture, development, and energy infrastructure consuming and fragmenting ecosystems faster than they can recover. This is happening now, at scale, with compounding effects on the biological systems civilization depends on, and it receives a fraction of the cultural and political attention that carbon does.
The irony is that the dominant political solution to climate — solar and wind at scale — directly accelerates the habitat problem. Solar panels cover ground. Wind turbines are sited in open landscapes, and the turbines, access roads, and transmission corridors fragment ecosystems in precisely the ways that drive species loss. More critically, wind farms placed in migratory corridors kill birds and insects at scale — raptors, songbirds, bats, monarch butterflies — disrupting population dynamics that span continents. The energy sources with the smallest physical and ecological footprint are nuclear and geothermal, both of which were legislated out of existence in the 1970s and systematically defunded since. A civilization running on nuclear and geothermal would produce the same energy on a fraction of the land, leaving the rest as habitat. The “green” energy buildout causes the biodiversity crisis it claims to solve. The genuinely low-impact energy sources are the ones the environmental movement spent fifty years fighting.
Twenty years of serious nuclear deployment is sufficient for the carbon timeline. The technology exists. The constraint is political, not physical — which means it’s the same kind of constraint that got circumvented when AI companies needed baseload power badly enough not to care about the politics.
Two Instruments
If energy is the one input everything else reduces to, the question writes itself: why not anchor money to it directly? The kWh token: one kilowatt-hour produced mints one coin. One kilowatt-hour consumed burns it.
The token isn’t a receipt for a specific unit of energy — it’s a claim on the pool. Energy is fungible: a kilowatt-hour is a kilowatt-hour regardless of whether it came from a nuclear plant in Tennessee or a solar array in Arizona. The boundary is the productive perimeter of the economy. Production crosses inbound — token minted. Consumption crosses outbound — token burned. Inside the boundary, tokens circulate freely against each other, interchangeable, because the energy they collectively represent is interchangeable. Verification happens at the border, not in the interior. This matters because it means the system doesn’t require tracking which energy unit backs which token — only that the books balance at the crossing points.
The first practical question is how this doesn’t immediately seize up: if coins only exist after energy is produced and vanish when energy is consumed, what do you actually transact with? The buffer solution: energy producers can pre-mint coins against their verified generation capacity for a fixed forward window — a month, a quarter, a year — posting 100% collateral for every coin issued in advance. As production is delivered and confirmed over that window, collateral releases proportionally. If a producer defaults, the escrow covers the gap. The monetary base at any given time is the aggregate of all forward commitments across all active windows: a nuclear plant committing to a year of output pre-mints a year of tokens today, and those tokens circulate for months before the underlying energy is delivered. Every coin in circulation is backed either by verified past production or by a fully-collateralized future commitment. No fractional reserve: a hundred kilowatt-hours of verified capacity supports a hundred coins, not a thousand. This puts the pre-minted token in a category that doesn’t quite exist in prior monetary systems: not backed by energy that exists yet, but by a verifiable commitment that it will, enforced by collateral that indisputably does. Harder than a fiat promise, softer than a commodity token, converging to real on a fixed schedule.
The design takes seriously a tension that Graeber’s history surfaces again and again: money is always being asked to do two incompatible things simultaneously. As a store of value, it should be scarce and appreciating — you want to hold it. As a transaction lubricant, it should be plentiful and stable — you want to spend it. Every monetary system in history has been trying to do both and failing at one side or the other. Commodity money deflates — England in 1696, the Ming Dynasty, and, in digital form, Bitcoin. Fiat extracts through Cantillon — the postwar dollar, Nixon’s improvisation, the petrodollar as pressure valve. Fractional reserve lending was the attempted square-peg solution: create enough credit to lubricate the economy while maintaining the fiction of commodity backing. It worked well enough to fund railroads and industrialization, and failed badly enough to produce the $846 trillion derivatives market and the debt cycle we’re now approaching the end of.
The kWh token separates the two jobs rather than trying to reconcile them. It’s explicitly a transaction lubricant — elastic, but not arbitrary. Supply grows with real energy production. Jevons runs in the right direction: cheaper energy makes more applications viable, which creates more demand for energy, which mints more tokens. The money supply grows with civilization’s productive frontier rather than with a central bank’s political calculations. For storing wealth, you need a different instrument — Bitcoin, or something with a hard cap, doing the one job it does well. Two instruments, each doing one job, neither trying to do both.
Critically, the kWh token is designed to depreciate against hard assets as civilization’s energy capacity expands. This is not a flaw. It’s the design. Cheaper energy means each token buys more productive activity — the good kind of deflation, driven by real efficiency gains rather than monetary scarcity. You are not supposed to save kWh tokens. Saving belongs in Bitcoin. The burn mechanic enforces this: you cannot hoard kWh tokens without cost, because they represent energy the economy needs to use. This is the structural inverse of fiat accumulation, where holding cash is free and the cost is externalized to everyone else through inflation.
Distributed production is the other key anti-capture property. Nuclear, geothermal, solar — the technologies now approaching the cheap energy threshold — can be generated anywhere. There is no Saudi Arabia of kilowatt-hours. Unlike oil, restricting energy supply hurts your own economy before anyone else’s — an OPEC for energy is self-defeating. The “printer” that mints new tokens is distributed across every generator on the grid. Proximity to the printer means proximity to actual energy production. You cannot extract without producing. A producer’s margin is the spread between their cost of generation and the market value of the tokens they mint — profit flows to whoever produces most efficiently, not to whoever sits closest to a central authority. The competitive pressure runs in exactly the direction you want: downward on energy costs.
The distribution goes all the way to the individual. A homeowner selling excess solar back to the grid is a minter. This is the most radical departure from every prior monetary system: in fiat, only central banks issue currency; in Bitcoin, only miners with expensive hardware mint economically; here, anyone who produces energy participates in monetary issuance proportionally. The long tail of individual prosumers adds monetary mass in aggregate and, crucially, makes the verification network harder to compromise — more producers means more crossover nodes reconciling more flows, and coordinating a consistent falsification across thousands of independent participants in dozens of jurisdictions becomes progressively less tractable. To shut down the monetary system you would need to shut down energy production itself. That’s not a governance attack surface. That’s physics.
This is also the structural answer to the capture question that the Bitcoin section raises. Bitcoin was captured at the governance layer — a small group of developers and mining pools could set protocol direction, and well-positioned interests could fund the people doing it. Energy production has no governance layer in that sense. A backed currency is a reference point: other currencies price against it, not the other way around. You can debase a fiat currency relative to an energy standard; you cannot debase the standard itself. And at the scale of 10x energy supply — production distributed across thousands of nuclear plants, geothermal installations, and solar arrays across every continent — no actor can corner or manipulate the supply without damaging their own economy first. The manipulation surface shrinks as the network grows. That’s the opposite of what happened with Bitcoin’s concentrated developer base and mining pools.
What the System Requires
There is a volatility problem with energy markets that any energy-backed currency must address honestly. Energy demand is highly inelastic: consumers have limited ability to substitute when supply tightens, so prices absorb the entire imbalance far out of proportion to the underlying supply shift. ERCOT spot prices during extreme weather events have reached hundreds of times their median — Winter Storm Uri in 2021 pushed prices to $9,000 per megawatt-hour against a typical $20-50. European electricity averaged below €40 per megawatt-hour for most of the previous decade, then hit over €200 average in 2022 on roughly a 10-15% supply shortfall — a 5x move on a modest disruption. A monetary system that inherits spot price volatility directly would reproduce the instability it was designed to replace.
The design requires two structural conditions to be stable. The first is overabundance — and specifically the right kind of overabundance. Solar and wind add capacity but don’t solve volatility; their output is intermittent, which is precisely why spot prices spike when weather turns. Nuclear and geothermal provide always-on baseload: a production floor that doesn’t fluctuate with the season, the time of day, or the wind. At 10x energy supply built on that foundation, disruptions consume surplus before they affect the monetary base. The second is temporal arbitrage through storage. Battery operators buy and store energy when supply is abundant and tokens are cheap, then discharge and sell when demand spikes and tokens are expensive. Their profit is the spread between those two states. This is its own economy — temporal arbitrageurs functioning like market makers — and it creates exactly the right investment incentive: build more storage, capture more spread, reduce volatility as a side effect. Critically, this is lending-like but physically constrained. A battery can only discharge what it charged. There is no fractional reserve equivalent for stored energy — you cannot lend kilowatt-hours you don’t have. The kWh token is not designed for a world of energy scarcity. In a world of energy abundance, the surplus itself becomes the stability mechanism, and the market that profits from managing that surplus builds the infrastructure to deepen it.
The fraud challenge is real, and it takes a specific form worth naming: not fake power plants, but signal fraud — hacked sensors, spoofed meter readings, falsified production data. Digital fraud, not physical fraud. The question is whether the network architecture makes that hard enough to matter.
The architecture distinguishes two types of nodes. Leaf nodes sit at individual producers and consumers — simple, unidirectional sensors confirming one flow direction each: energy in, or energy out. Crossover nodes sit at junctions where multiple channels converge, reconciling all inflows against all outflows. The analogy is cardiovascular: a heart can detect a hemorrhage by observing that more blood left than arrived, without trusting every capillary individually. Directional channel topology — dedicated inbound and outbound flows at each junction rather than bidirectional pipes — converts the verification problem from “is every meter in the entire network honest?” (intractable) to “do flows balance at each junction?” (auditable). A crossover node that sees 100 kWh arriving but 94 kWh departing knows exactly which segment to investigate. You cannot spoof a reading at one leaf node without the imbalance propagating to every downstream crossover node. The books must balance across the whole network. To cheat without detection requires coordinating a consistent falsification across many sensors simultaneously — the attack surface scales with the network’s density, not with the attacker’s ambition. Compare this to the Cantillon machine, where proximity to a money printer yields purchasing power for zero productive contribution. The incentive gradient runs in the opposite direction.
The hard problem sits one level down from what the topology solves. A compromised leaf node gets flagged by crossover reconciliation. But a compromised crossover node itself, a firmware modification that makes genuine hardware report false readings, or a man-in-the-middle between the physical sensor and the network data feed — these are signal-layer attacks that flow-balancing architecture alone doesn’t address. The real verification challenge isn’t thermodynamic: it’s tamper-resistant metering at the hardware and firmware layer. That’s an active research area in smart grid and IoT security. It isn’t solved. But there’s reason for optimism about where the baseline sits: modern grids already perform flow reconciliation routinely. Utilities detect energy theft through what they call non-technical loss monitoring — if 100 kWh flows into a distribution zone and only 94 kWh flows out through metered endpoints, the discrepancy flags automatically. The infrastructure for detecting imbalances at scale already exists and works. The unsolved problem is hardening the signal layer against deliberate, coordinated manipulation — active adversaries rather than the passive discrepancies the grid already catches. That’s a harder problem than the one already solved, not a different category entirely.
This is exactly where Burke’s touchstone logic applies. Commodity coinage only became viable once the touchstone existed. Every monetary leap has required a new verification technology. The touchstone for precious metals. Cryptographic proof-of-work for Bitcoin. Tamper-resistant metering at the signal layer for energy — the same category of problem, one level harder. I’m describing a direction, not a blueprint. The whitepaper has a full section on the mechanism design.
Nixon closed the gold window in 1971 and improvised an energy standard — the dollar backed not by gold in a vault but by control of oil flows, administered through aircraft carriers and cartel agreements rather than through protocol. We’ve been on an energy standard for fifty years. We just don’t call it that, because the version we have was built by accident and is run by coercion.
That improvisation is ending. The underlying debt cycle is in its terminal phase. No sovereign candidate can replace the dollar without recreating the same problem under new management. The petrodollar’s demand prop is eroding. The derivatives market at eight times world GDP is what fractional reserve lending looks like at civilizational scale, and it has no soft landing.
The 21st century has something that previous monetary systems lacked: distributed energy production, cryptographic verification, and thermodynamic constraints that no legislature can override and no central bank can erode. These are the components of a monetary substrate that doesn’t reward proximity to a money printer — that makes extraction more expensive than production not through rule of law, which can be dissolved in an afternoon on television, but through physics.
Every monetary substrate is a theory about what has value. Fiat says: whatever the state declares. Gold says: whatever geology made scarce. The petrodollar said: whoever controls the energy supply. The kWh token says energy itself — the one input that everything else reduces to, produced by anyone with the technology to generate it, verified by physics rather than by trust.
The question is whether we build the successor deliberately or discover what replaces the dollar the hard way.
</iXi>
The full mechanism design for the kWh token — including the buffer architecture, verification framework, and governance implications — is developed in a whitepaper on how communities and civilization might build more resilient institutions and better governance. More there for anyone who wants to go deeper.
Open questions
These aren’t objections to the design — they’re the places where the design is genuinely unresolved. Listing them here because honest accounting of what’s unsolved is more useful than pretending it isn’t.
Protocol governance. Who controls the validation rules — what counts as verified production, what collateral qualifies, how crossover nodes are certified? If that’s a foundation or a consortium of large utilities, it’s a new governance attack surface. Different from Bitcoin’s developer concentration problem, but the same category of risk. An open specification (like HTTP — a standard with many independent implementations) may be more durable than a single protocol authority.
Grid-level censorship. Individual producers can route production to a pseudonymous wallet — the leaf node confirms energy crossed the boundary and mints to an address without needing to know who owns it. The long tail of solar producers becomes a population of effectively anonymous minters, difficult to censor at scale. Large future-minters posting escrow collateral require more identification (counterparty risk in the collateral process demands it), but there are far fewer of them, they’re already regulated entities, and censoring critical energy infrastructure carries political costs that censoring a software wallet doesn’t. The remaining open question is whether governments can compel grid operators to correlate wallet addresses to meter IDs — technically feasible if the grid operator has that mapping, an open design question whether they need to.
Regional tokens vs. global standard. Energy prices vary dramatically across geographies. A Texas kWh and a Japanese kWh represent the same physical quantity but different economic values. Multiple regional tokens on the same physical grid create a double-minting risk (the crossover architecture detects this as flow imbalance), but regionally distinct protocols on separate grids could coexist cleanly, interchangeable through exchange. Whether one global standard or a family of regional tokens is the right architecture is an open question. Nuclear expansion probably reduces this delta.
Protocol migration and lower lock-in. Because the token is a claim on a real good rather than the good itself, migrating from one protocol to another is updating the accounting layer, not convincing monetary mass to move. This is meaningfully less lock-in than Bitcoin, where the value IS the network consensus. Whether this is a feature (easier to fix governance failures) or a risk (easier for states to force migration to a compliant protocol) depends on implementation.
The bootstrapping window. The stability arguments — overabundant supply, mature storage markets, distributed production — describe a world that cheap energy creates, not the world we’re in now. A kWh token system deployed into today’s energy markets inherits today’s volatility. The transition period between here and there is a real design problem that the whitepaper addresses but this article does not.
Responses, objections, and extensions welcome.