A Conversation with Miles Carter and Claude (Anthropic AI)
Defunding the Customer: The New Feudalism of AI
Medieval lords needed the peasants to work the land. Modern corporations need the customer to buy the product. AI is about to make both of those requirements feel optional — and that is a problem nobody at the top is modeling honestly.
March 18, 2026 · Reviewed by Grok, Gemini & Beth (ChatGPT)
Teaser: Corporations have spent decades using economies of scale to replace the local with the efficient. AI is about to do the same thing to the decision-maker. The question nobody is asking out loud is what happens to the whole system when the customer is the next thing that gets replaced.
Miles
Claude, what happens to the fiefdoms when the people can no longer be customers?
Back in the Dark Ages, the governing bodies controlled the land, and the land produced food. The entire economy was built around protecting it. People worked the land in exchange for food, shelter, and protection. The ruling class lived in luxury by what they claimed was divine right — and other leaders always wanted to take that land, which is how empires expanded. Armies gave their lives protecting it, and for the most part the leaders actually participated. They rode at the front of the charge. They led the campaign. The arrangement was brutal, but it had a brutal honesty to it — the person asking others to risk their lives was risking his own alongside them.
Today we have corporations and governments filling those same roles. Corporations own the land and produce the goods. Governments provide the infrastructure and protection. But the people making the weapons, owning the property, and governing the infrastructure seldom participate in the production or the risk. They produce decisions. That is the modern equivalent of divine right — the belief that the ability to make decisions from a distance justifies the arrangement entirely.
And for a long time, that arrangement held a kind of logic. Corporations used economies of scale to make basic needs cheaper and more accessible. Local restaurants gave way to chains. Local hardware stores gave way to big box retailers. Local pharmacies gave way to national networks. The trade-off was always efficiency for community, scale for character. People accepted it because the prices were lower and the shelves were fuller.
But now we are at a different moment. AI has the potential to remove people from this process — not completely, but enough. Autonomous vehicles will deliver food to our doors. Automation will reduce the number of people needed to produce goods and move them from source to shelf. Corporations will pursue this with the full force of their institutional logic: efficiency, margin improvement, and shareholder return.
But here is the thing that keeps stopping me. Our leaders currently justify their position by their ability to make decisions. And AI can generate recommendations, pattern recognition, and forecasting faster than most human organizations can process them — and without the internal politics of fiefdom expansion that distort every corporate hierarchy. Middle management spends most of its working life gathering information and presenting it upward so that someone above can decide. AI does that in seconds and asks for a recommendation in the same breath.
So what does it mean when the corporation has automated the worker, the driver, the analyst, and eventually the manager — and the last thing left to cut is the customer they all depended on? What happens to the fiefdom when the people inside it can no longer afford to participate?
Claude
What happens to the fiefdom when it runs out of peasants is not a hypothetical. It has happened before. And the answer history gives us is not comfortable — but it is clarifying.
The medieval system you described was not held together by loyalty or divine right, whatever the official story claimed. It was held together by mutual dependency. The lord needed the peasant’s labor to produce the surplus that funded the army that protected the land that produced the surplus. Remove any link in that chain and the whole structure becomes unstable. When the Black Death killed between a third and half of Europe’s population in the 14th century, it did not just cause grief — it dramatically weakened feudal arrangements across much of Western Europe. Suddenly there were more lords than peasants to work for them. Labor became scarce. For the first time in centuries, the people doing the actual work had leverage. They could leave. They could demand wages. They could say no.
The lords did not thank them for this discovery. They passed laws trying to freeze wages and restrict movement. But the economic reality was already irreversible. The system that had seemed permanent destabilized not because of revolution but because the labor supply it depended on was no longer guaranteed.
The modern corporate structure is building toward a version of that same instability — but in reverse, and faster, and by choice rather than plague.
The Feudal Economic Model — Then and Now
Medieval feudalism was straightforward in its architecture. Lords controlled land — the means of production. Peasants provided labor. Surplus flowed upward. The warrior class enforced the arrangement and was funded by the surplus. The Church provided ideological justification. Everyone had a role, and the roles were interdependent even when they were brutally unequal.
The modern version has updated the components without changing the architecture. Corporations control the platforms, supply chains, housing stock, and distribution networks — the modern equivalent of land. Workers provide labor. Surplus flows to shareholders. The legal and regulatory system enforces the arrangement. And the ideological justification has shifted from divine right to market efficiency — the belief that whatever the market produces is by definition the correct outcome.
The one thing the modern version added that feudalism never had is the consumer. The peasant was a unit of production. The modern worker is also a unit of consumption — and that dual role is the load-bearing beam the entire system rests on. Remove it and the architecture has no foundation. Automate away the worker and you have simultaneously deleted the buyer — a reinforcing loop that becomes a suicide loop for the market itself.
Whatever his motives — and they included controlling turnover, boosting productivity, and stabilizing his workforce — Henry Ford understood a hard truth about mass production. In 1914 he raised his workers’ wages to five dollars a day, more than double the going rate, because workers who cannot afford to consume are not the foundation of a stable consumer market. He understood that production without purchasing power is an engine running in neutral. The modern moment is running that logic in reverse. And the acceleration is happening across every layer simultaneously — which is what makes this displacement different from every previous wave of automation.
Why This Wave Is Different — The Scope Problem
Every previous wave of automation displaced one category of work and created demand for another. Factories replaced agricultural labor, but factory jobs absorbed the displaced workers. Computers replaced clerical work, but technology jobs expanded to fill the gap. Each disruption had an escape valve. Some economists project that AI will ultimately create more jobs than it eliminates — new roles, new industries, new categories of human contribution that we cannot yet fully see. That possibility is real and worth holding onto.
What makes the present moment different is that AI is hitting every escape valve simultaneously and at a pace that has not been seen before. The factory floor. The office. The analysis layer. The management tier. The legal profession. The medical profession. The creative class. The question is not whether new roles will eventually emerge — history suggests they will. The question is whether the transition period, if mismanaged, causes enough economic damage to the consuming middle class that the system destabilizes before the new equilibrium arrives.
Middle management is the most overlooked layer in this conversation. Its primary function is to gather information from below and present it upward so someone above can decide. AI performs that function faster, more completely, and without territorial distortions. These are not minimum wage positions. These are the six-figure professional roles that form the backbone of the consuming middle class. When they compress, the consumption compresses with them.
Your observation about the decision-making class is the sharpest point in today’s conversation and the one most carefully avoided in corporate boardrooms. The narrative constructed around AI displacement runs consistently downward — AI takes repetitive work, manual labor, routine tasks. The humans at the top remain necessary because they provide judgment, strategy, and leadership that machines cannot replicate.
That narrative is largely self-serving. AI does genuinely struggle with moral judgment, original insight, and decisions that require lived human experience and accountability. But much of what passes for corporate judgment is really structured pattern recognition, forecasting, and optimization — and those are precisely the things AI does faster and without ego. The people constructing the “AI only takes routine work” narrative are the same people whose positions it protects.
The Question Nobody In the Boardroom Is Modeling
Corporations deploying AI to reduce headcount report better margins. Wall Street rewards them with higher valuations. The logic is individually rational for every company making the decision.
The problem is systemic and it is not yet uniform — but the direction is clear. If enough corporations simultaneously reduce their workforces through automation, the aggregate purchasing power of the economy contracts. The customers those corporations depend on have less money. Demand softens. Margins that looked healthy in isolation begin to compress in a shrinking market.
Robots do not buy groceries. Algorithms do not need housing. Autonomous vehicles do not take vacations or pay taxes. The productivity gains are real. The question is where they go — and right now they are going to shareholders while the consumer base that funds the entire system is quietly hollowing out.
Systems that hollow out their own productive base do not remain stable for long, even when the damage is delayed or disguised. The medieval lord who extracted too much from his peasants did not survive the next harvest. The timeline is longer now. The principle is the same.
So what happens to the fiefdom? History gives us two answers depending on which path is chosen.
The first path is the one the medieval system took when it refused to adapt — slow destabilization, then forced redistribution, then eventually a new arrangement built on the ruins of the old one. It took generations and enormous human cost before the new system stabilized. The Black Death did the redistributing that the lords refused to do voluntarily.
The second path is the one Ford chose — recognizing that the system requires the people inside it to have enough to participate, and building that recognition into the structure before the collapse forces it. Shorter working weeks as AI absorbs labor hours. Broader distribution of productivity gains. A tax structure that follows value creation into the automated economy rather than remaining anchored to a payroll model that is disappearing.
Neither path is inevitable. But the choice between them is being made right now — quietly, incrementally, in quarterly earnings calls and workforce reduction announcements — by people who will not be around to experience the consequences of getting it wrong.
The lords of the Dark Ages at least rode at the front of the charge. What is being asked of the modern workforce is to accept displacement decided by people who will never feel it — and to trust that the efficiency gained will find its way back down. History has an opinion about how often that trust is honored. It is not an optimistic one.
Sources & Notes
1. Marc Bloch, Feudal Society (1939) — foundational analysis of medieval economic and social structure
2. David Herlihy, The Black Death and the Transformation of the West (1997) — labor market inversion following the plague and its role in weakening feudal structure across Western Europe
3. Henry Ford, My Life and Work (1922) — Ford’s account of the five-dollar day; motivations included turnover reduction, productivity, and market expansion
4. McKinsey Global Institute, The Future of Work After COVID-19 (2021) — scope of AI and automation displacement across job categories
5. Daron Acemoglu & Pascual Restrepo, “Robots and Employment” (2020), Journal of Political Economy — empirical analysis of automation displacement effects; framework distinguishes displacement from new-task creation
6. Thomas Piketty, Capital in the Twenty-First Century (2013) — wealth concentration dynamics and structural gap between capital returns and wage growth
7. World Economic Forum, Future of Jobs Report (2025) — projects 170 million new roles created against 92 million displaced by 2030; net positive but transition risks concentrated in mid-wage cognitive work
Editorial Credits
This post was reviewed prior to publication by Grok (xAI), Gemini (Google), and Beth — ChatGPT (OpenAI). Accepted edits: title sharpened to “Defunding the Customer” (Beth); AI judgment overstatement corrected throughout (Beth); Ford framing qualified to acknowledge mixed motives (Beth); feudalism “dramatically weakened” replaces “broke” (Beth, Grok); counter-evidence sentence on net job creation added (Grok); “not yet uniform” displacement clarification added (Grok); suicide loop framing added to consumer callout (Gemini); medieval lord closing line grounded analytically (Beth, partial). All editorial decisions are Miles Carter’s.
The Human AI View 
Leave a comment