The Bubble of Babel

The Bubble Expert

I was watching Jeremy Grantham on the Diary of a CEO. I make a practice of studying arguments against my own position. I lean libertarian. I suspect the state and its motives. Grantham sits on the opposite side. He wants more regulation, more taxation, more state intervention in markets and the environment. But his diagnostic work is worth taking seriously, because his bubble thesis is pistomechanical whether he knows it or not.

Grantham understands that markets run on belief. Prices are psychology, not mathematics. The same real innovation that justifies real investment also triggers a belief overshoot that prices cannot sustain. Railroads were real. The railroad bubble was not. The internet was real. The dot-com valuation was not. He has documented this pattern across every major asset class in modern history, and the pattern is always the same: a real thing generates real results, the results justify belief, the belief overshoots the reality, and the correction destroys the people who mistook the trance for the territory.

He even sees the incentive structure that keeps the trance running. Investment advisors will never tell you to exit. Their business model requires your participation. They are paid to keep you inside the bubble, and the longer the bubble runs, the more reasonable it looks, until the people warning you seem like the irrational ones. That is clean pistomechanical observation. Grantham is reading the code.

But as I watched, something crystallised. Every time he reached for a solution, the solution was the same node. The state should regulate chemicals. The state should restructure capitalism to be family-friendly. The state should tax corporations into better behaviour. Every prescription assumed that the state is a trustworthy operator. That if you could just get the right people running it, the right policies installed, the machinery would work. He never audited that assumption. He could see belief mechanics operating in every market, every cycle of euphoria and collapse. He could not see the same mechanics operating in his own foundational commitment. The state sat below his diagnostic threshold.

Then a question formed that I could not put down. What if the state itself is a bubble?

Grantham’s career rests on a single observation: that human beings mistake overextended belief for permanent reality. The nation-state is roughly four hundred years old. The state as a concept, centralised authority governing a population within a territory, is perhaps six thousand years old, dating to the first Mesopotamian city-states. It has generated real results. Coordination at scale, legal systems, infrastructure, military defence. The results justified belief. The belief deepened until it stopped being experienced as belief and became the ground everyone stands on. If Grantham is right that the discipline is to identify when belief has overshot reality and prepare before the correction arrives, then by his own methodology he should be asking whether the institution he trusts most has itself overextended. He is preparing for every bubble except the one he is standing inside.

Grantham is not wrong about the bubbles he can see. He is brilliant about them. The pistomechanical observation is that his analysis is incomplete. He never turns the diagnostic tool on the layer he is standing on. And that is what this essay does.

The Gradient

Keynes believed markets are unstable. Left alone, they overshoot, crash, and destroy livelihoods. The state must intervene: spend during panics, cool during booms, employ when employers will not. Friedman said leave them alone. Markets self-correct. State intervention causes more distortion than it cures. Minimise the bureaucracy. Let prices carry information.

These two positions have divided Western political economy for a century. Every election, every budget debate, every argument about taxes or regulation or welfare reduces to some gradient between them. Left versus right. More state versus less state.

Both positions accept the same substrate. Both treat the nation-state as the non-negotiable unit of political organisation. They argue about settings. Neither questions the machine.

The entire political spectrum runs on a single variable: how much do you trust the state?

The far left trusts it completely. Tax, spend, regulate, nationalise, redistribute. The state knows better than the market. The centre-left trusts it with guardrails. Markets can work, but they need oversight. Regulation keeps capitalism honest. The centre-right trusts the market more. Lower taxes, fewer rules, smaller government. The state is necessary but should be restrained. The libertarian trusts the state almost not at all. Enforce contracts. Defend borders. Everything else is interference.

Move along the spectrum and the dosage changes. The prescription stays the same: the state is the only legitimate container for political life. Nobody on this spectrum proposes an alternative container. The most radical libertarian still believes in the minimalist state. Keynes and the anarcho-capitalist disagree about everything except the one thing that would make their disagreement interesting.

The Installation

The nation-state is roughly four hundred years old. Date it from the Treaty of Westphalia in 1648, when European powers agreed that sovereignty belongs to territorial governments rather than the Church, the Emperor, or dynastic claims. In its modern bureaucratic form it is closer to two hundred. Before that, political life ran on dynasties, tribes, city-states, religious networks, feudal hierarchies, and extended family alliances. The territorial bureaucratic state did not exist.

People owed loyalty to a lord, a clan, a faith, or a guild. The idea that every person on earth should be a citizen of exactly one government, carrying documents issued by that government, taxed by that government, and conscriptable by that government, would have struck a medieval European or an ancient Israelite as bizarre.

The state replaced these older structures for a specific reason. Industrialisation and mass warfare required resource extraction at a scale that family networks and city-states could not achieve. Fielding a modern army, building a railway, administering a colonial territory demanded centralised taxation, standardised law, mandatory schooling, and the administrative machinery to enforce all three. The state was a compression. An emergency adaptation to a new environmental pressure. Exactly like the compressions mapped in the Lossy Compression Cascade: a survival move that gets filed as a permanent feature.

Once installed, the state did what every deep-layer belief does: it made itself invisible. The fish does not see the water. Most people alive today cannot imagine political life without a state. They can imagine a different state. Democratic instead of authoritarian. Capitalist instead of communist. Secular instead of religious. They cannot imagine no state at all. The state has descended below the opinion layer, below the value layer, below the identity layer, into the perceptual firmware that shapes how reality is experienced before conscious evaluation begins. It is the deepest secular trance currently running on the majority of humans alive.

And the state maintains its installation the way any deep-layer belief system does: through daily ritual. The pledge of allegiance is a morning affirmation recited by children in a light trance of social compliance. Hand on heart, eyes on flag, voice in unison. The national anthem before a sporting event is a collective effervescence ritual. The passport is a portable identity document that tells you who you belong to before you tell it where you want to go. The tax return is an annual confession of your entire economic life to a single authority. The currency in your pocket carries the state’s symbols and demands your trust with every transaction. These are the same mechanisms that keep a religious installation running: repetition, communal synchronisation, identity documents, confession, sacred symbols. The difference is that the religious practitioner usually knows they are practising. The citizen usually does not.

The Salvific Pattern

Every centralising movement follows the same sequence. There is a real problem: suffering, disorder, inequality, scarcity. A figure or institution identifies the problem. So far, legitimate. Then comes the salvific move: this mechanism will solve it. The Party will deliver justice. The Market will allocate optimally. The Dollar will store value. The Algorithm will know you better than you know yourself. The State will protect you.

Each asks for the same transfer: place your belief in this specific node, and it will manage reality on your behalf.

And the transfer works. That is what makes it dangerous. Communism did industrialise Russia in a single generation. The dollar did function as a reliable store of value for decades. Capitalism did generate wealth at a scale no prior system matched. The results become the proof that justifies deeper belief, which justifies more centralisation, which justifies more dependency, until the system is load-bearing on a single node that was never designed to carry that weight. Then the node fails, and the people who built their lives around it have no fallback.

This is the pattern in every bubble, every political collapse, every failed utopia. A real innovation justifies real belief. The belief overshoots the reality. The overshoot gets mistaken for a feature of the innovation rather than a property of the belief. Then the correction arrives and destroys the people who could not distinguish between the tool and the idol.

Idolatry as an Engineering Problem

Idolatry, read mechanically, is the error of placing ultimate trust in a specific mechanism.

The golden calf was not worshipped because it was a cow. It was worshipped because it was specific, visible, and controllable. Moses was gone. The mountain was terrifying. The unknown was unbearable. So the people built a node and centralised their belief around it. The text treats this as the prototypical mistake. The problem was not the material. The problem was the architecture: centralising trust in any specific, controllable mechanism is the move that makes a system brittle.

Reality is emergent. It generates novelty that no model anticipated. The market cannot predict itself. The state cannot administer what it cannot model. The algorithm optimises for its training data, which is the past. Every centralising mechanism is a bet that the future will resemble the conditions under which the mechanism was designed. That bet pays off for a while. It always stops paying off, because reality does not repeat. It generates.

Centralisation, then, is an irrational belief in control. It assumes that a specific node can track an emergent process. The assumption holds locally and temporarily. Over sufficient time and sufficient novelty, it always breaks. And the more completely a population has centralised its trust in that node, the more catastrophic the break.

The Anti-Centralisation Clause

If the vulnerability is biological and permanent, no technology can fix it. The pressure to centralise lives in the dominance-hierarchy firmware that predates language, predates the human species, predates mammals. De Waal documented it in chimpanzees. Lorenz documented it in geese. Give humans a decentralised substrate and they will re-centralise around it, because the instinct to organise around a single dominant node is older than the instinct to speak.

The internet was supposed to decentralise information. It produced Google, Facebook, and Amazon. Bitcoin was supposed to eliminate trusted intermediaries. It produced mining pools, exchange dependency, and a community that watches two men’s social media posts to decide whether to buy or sell. Each technology delivered real decentralisation and then watched its users re-centralise around new nodes, because the technology addressed the protocol layer while the centralisation pressure lives in the biological layer.

The problem cannot be solved by choosing the right technology. It can only be managed by a cultural norm that flags every centralisation as a concession, never a feature. A constitutional clause. An immune response against single points of failure.

This clause already exists. It has been running for over three thousand years.

The Torah begins the story of Israel with an anti-centralisation narrative. Babel is a centralisation project. One language, one city, one tower, one name. The text presents it as a civilisational achievement. Then it breaks it. Not by destroying the tower but by scattering the builders. Forcing diversity back into the system.

The first Israelite is called to leave. Abraham is Ivri, from the root avar: one who crosses over. The identity is constituted by departure from the centralised architecture. Before any law is given, before any land is promised, the founding act is a crossing-out. A refusal to remain inside Babel.

At Sinai the principle gets codified. You shall have no other gods. Read mechanically, this is a constitutional clause. It does not argue that other gods are false. It says: even if they deliver results, do not centralise your trust in them. The clause operates at the structural level. It asks one question: does this concentrate authority in a single point? If yes, reject it. Regardless of the results.

The Immune Response

This clause produced a measurable historical effect. Every time a salvific centralisation emerged from within Jewish civilisation, the clause triggered.

Christianity offered a centralised mediator between the individual and God. Judaism rejected it. Sabbateanism offered messianic fulfilment through a single charismatic figure. Rejected. Frankism offered antinomian liberation through a single leader who claimed to have transcended the law. Rejected. Each movement addressed a real need. Each attracted serious people for serious reasons. Each was identified by the anti-idolatry clause as a centralising installation that would collapse the distributed architecture, and expelled.

The clause does not need to prove these movements false. It identifies them as structurally dangerous. A system with a single mediator, a single messiah, a single charismatic authority, is a system with a single point of failure. The immune response is not theological. It is architectural. And it has kept a distributed civilisation intact across dozens of catastrophic disruptions that destroyed every centralised civilisation they touched.

The Political Implication

If the anti-idolatry clause is an engineering principle rather than a religious preference, it applies to secular institutions with the same force it applies to golden calves.

The state is a tool. Taxation, courts, infrastructure, defence: useful mechanisms for organising cooperation at scale. The state becomes an idol the moment it is treated as the only legitimate container for political life. The moment a person cannot imagine organising their family’s safety, wealth, education, or identity without the state, the tool has become an idol. They are worshipping it in the precise mechanical sense: they have centralised their trust in a specific node and cannot conceive of operating without it.

The market is no different. Price signals, voluntary exchange, competition: useful mechanisms. The market becomes an idol when someone believes it will always self-correct. “The market will sort it out” is a faith statement. It asks you to trust a mechanism to manage a future that the mechanism cannot model.

Technology follows the same logic. Smart contracts, decentralised ledgers, cryptographic verification: useful mechanisms. Blockchain becomes an idol when someone claims it has solved the coordination problem forever. No protocol solves that problem forever. The coordination problem is permanent, because the biological pressure toward centralisation is permanent.

Keynes worshipped the state. Friedman worshipped the market. Both believed a specific mechanism could be trusted to manage emergent complexity. Both were wrong at the level of architecture, regardless of which was right at the level of policy. The debate between them is an argument about which idol to worship. The pistomechanical question is whether to worship idols at all.

The Bubble of Babel

The distinction is operational.

A tool is something you use and put down. You assess its performance against the task. When it stops working, you pick up a different tool. You hold no loyalty to it. You do not build your identity around it. You do not assume it will work tomorrow because it worked today.

An idol is something you cannot put down. You have installed it below the level where assessment operates. It has become part of your rendering engine. You experience its outputs as reality rather than as one mechanism’s interpretation of reality. When it fails, you blame the failure on something other than the idol, because questioning the idol would require you to question the layer of your operating system that you experience as the ground you stand on.

The free person is the one who uses tools without worshipping them. Who holds operational trust in specific mechanisms while refusing to install any of them as ultimate. Who holds multiple jurisdictions, multiple currencies, multiple allegiances, multiple frameworks, and treats each as a tool that may need to be put down when the conditions that made it useful change.

But the free person is not a nihilist. The anti-idolatry clause does not say “trust nothing.” It says “no other gods before Me.” There is a positive referent. Read mechanically: trust the generative, emergent process that produces reality. The process that no node can model, no state can administer, no algorithm can predict. The free person does not merely refuse to centralise. They hold active trust in the ungovernable process that generates novelty. The cynic says nothing is trustworthy. The postmodernist says all frameworks are equivalent. The Ivri says one thing is trustworthy: the process itself, the emergent unfolding that no mechanism can contain. Everything that asks to be installed in its place is an idol.

The word for this posture, in the oldest tradition that practised it, is emunah. Usually translated as “faith.” More accurately: the active trust in what cannot be seen, modelled, or controlled. The willingness to cross over from Babel without a map. And to keep crossing.