A Manifesto for the Age of Destabilization (v5)
Revised April 2026. The previous version of this manifesto was written hours ago. It is already out of date. That fact is itself the argument.
I.
Stop planning for next year. Next year is a meaningless concept now.
The bargain that organized human civilization for three centuries — sell your labor, earn your place, consume to keep the engine running — did not slowly erode. It broke. It broke in months. And it is about to break again, harder, before you have finished understanding the first fracture.
In late 2025, artificial intelligence crossed the threshold from human assistant to human replacement. Not in a laboratory. Not in a research paper. In production. In companies. In the real economy. Entire categories of knowledge work — creative, analytical, administrative, financial, legal, managerial — became automatable not in theory but in practice. Companies didn't pilot AI. They fired humans and replaced them with API calls.
That was ninety days ago.
Now a new generation of AI models is already in testing. Models described by their own creators as a step change beyond the systems that already triggered mass displacement. Not incrementally better. Qualitatively different. More capable in reasoning, coding, analysis, and autonomous operation by an order of magnitude.
But capability is only half the story. Something has changed in who can deploy that capability and how fast.
A single person with AI can now build in weeks what previously required teams of hundreds and years of development. Industry-transforming products — systems sophisticated enough to restructure the labor model of an entire global sector — are being created by one or two people in the time it used to take to write a business plan. What took 350 engineers and four years in 2022 now takes a handful of people and six weeks.
This is not a metaphor. It is happening now, across every industry, simultaneously. The barriers that once slowed disruption — capital requirements, engineering talent, development time — have collapsed. Any person with domain expertise and access to current AI can build a product that restructures the economics of their entire field. The tools of industry-scale transformation are now available to anyone, deployable in weeks, and improving with every model generation.
That means disruption is no longer something that happens to industries one at a time, driven by well-funded startups over multi-year cycles. It is happening to all industries at once, driven by individuals, at a pace measured in weeks. The number of people capable of fundamentally reshaping a sector's labor model has gone from hundreds to millions. And each one of them is building on AI that is itself getting dramatically more powerful every few months.
This is not a forecast. This is a pattern already in motion.
If you are reading this and feel like the ground is still solid beneath you, it is not. You are standing on momentum. Momentum is not the same as ground.
II.
What is happening is not a single crisis. It is four simultaneous fractures, each one accelerating the others, forming a self-reinforcing loop that no individual actor can stop. The loop is turning faster than anyone predicted even weeks ago.
THE FIRST BREAK IS LABOR.
AI is not automating tasks. It is eliminating roles. Not over years. Over quarters.
In the first ninety days of 2026 alone, nearly 250,000 jobs have been publicly eliminated — almost 2,750 per day. And the pace is accelerating. When a $420 billion company fires ten thousand senior engineers by email to reassure investors, the permission structure for every other company changes overnight.
But the layoffs are only half the collapse. The other half is silent.
Companies are not hiring. Roles that would have been created are never opened. Growth that would have required headcount is being routed through AI instead. A company that would have hired fifty new graduates this year hires five and gives them AI tools. Forty-five jobs that will never exist. No layoff announcement. No headline. Just absence.
New graduate hiring has already declined by twenty-three to thirty percent. That is a generation being locked out of the labor market at the entry point. Not fired. Never let in. The entire pipeline of human capital development — internship, junior role, skill building, career progression — is being severed before it begins.
This changes the math of collapse. Unemployment measures how many people fell out of the boat. Hire rate measures how many will never get in. Both numbers are moving in the wrong direction simultaneously. The economy is draining from two openings at once — workers leaving through layoffs while new workers are blocked from entering. The pool empties faster than either force alone would predict.
And the two forces feed each other. Fewer new hires means fewer new consumers with spending power. Less spending means less revenue. Less revenue means more cuts to existing staff. More cuts means more unemployment. More unemployment means less demand. Less demand means even less reason to hire. The loop has two intake valves now, not one.
This also breaks the last counterargument with any weight — that displaced workers would be reabsorbed as the economy grew. Growth is no longer creating jobs. Revenue can increase while headcount decreases. Growth and employment have decoupled. The economic models that assume recovery equals hiring are built for a world that no longer exists.
Previous technological disruptions displaced workers from one sector into another over decades. This displacement is hitting every knowledge-work sector simultaneously and the timeline is months. A worker laid off from a SaaS company cannot retrain into marketing because marketing is being compressed. Cannot pivot to legal analysis because legal is being compressed. Cannot fall back on content creation because content was compressed first. And even if they could — the roles they'd retrain into are not being filled with humans.
And the displacement frontier is no longer advancing linearly. It is being pushed outward by millions of individuals simultaneously building AI-native products that restructure entire industries from the inside. A single founder can now build a system that halves the labor requirements of a global sector — in weeks, not years, with no venture funding, no engineering team, no infrastructure beyond a laptop and an AI subscription.
This changes the math of disruption entirely. Previously, industry transformation required concentrated capital and coordinated effort. It was slow because it was expensive. Now it is fast because it is cheap. The cost of building something that restructures an industry's labor model has fallen to near zero. That means every industry with inefficient human labor — which is every industry — is simultaneously being targeted by an army of builders who can see the inefficiency and now have the tools to eliminate it.
The service economy was supposed to be the buffer. Hospitality, fitness, food service, retail, personal care — roles that require physical presence and human interaction. These jobs were considered safe from AI displacement. They are not safe from AI-augmented displacement, where a single system can make one worker as effective as three by handling the cognitive overhead — memory, timing, situational awareness, decision-making — that previously required experience and headcount. The service economy doesn't need to be automated. It needs to be compressed. And the tools to compress it are being built right now, by individuals, in weeks.
The robotics frontier will extend this further into physical labor. But that timeline is irrelevant to the immediate crisis. The knowledge economy and the service economy are being hollowed out simultaneously — one by AI replacing humans, the other by AI making fewer humans sufficient. Both are happening now.
THE SECOND BREAK IS DEMAND.
Every company replacing humans with AI is destroying its own customers. Every individual building an AI product that compresses an industry's workforce is removing consumers from the economy. This is the paradox that will collapse the system and it is already in motion.
Workers are consumers. Wages are spending power. Spending power is revenue. Revenue is the reason companies exist.
The loop is exponential and it is turning now. Labor costs are cut. Consumer spending contracts. Revenue declines. Boards respond with further automation. More labor is removed. Spending contracts further. Each quarter is worse than the last not by a fixed amount but by a growing multiple.
And the loop is being fed from both sides. On one side, layoffs remove existing consumers from the economy. On the other, hiring freezes prevent new ones from entering. The demand base is shrinking from above and being blocked from replenishing from below. Previous economic models assumed that contraction on one side would be offset by growth on the other. That assumption is dead. Both valves are closing simultaneously.
But the demand destruction is not limited to the direct employer-employee relationship. This is where the full scale of the crisis becomes visible.
Every displaced knowledge worker is not just a customer of the company that fired them. They are a customer of the entire service economy. The coffee shop near their office. The gym they belonged to. The restaurant where they ate lunch. The dry cleaner, the barber, the daycare center. The weekend trips. The streaming subscriptions. The birthday dinners.
These businesses may never face direct AI labor compression. A barista's job is not threatened by a language model. But a barista's job is absolutely threatened by the disappearance of the customers who bought the coffee. The service economy does not need to be automated to be destroyed. It only needs to lose its demand base.
And this is where the network effects compound the crisis exponentially. When a corporation displaces a thousand knowledge workers, the immediate loss is their spending at that company. The secondary loss radiates into every service business those workers patronized. Those businesses lose revenue. They cut staff or close. Their former employees stop spending. That loss radiates into yet another ring of businesses. Third-order effects hit the landlords, the suppliers, the local tax base.
The speed differential is critical. Corporate AI displacement happens in planned waves — quarterly restructurings, strategic reductions. Service economy fallout will be faster and more chaotic because small businesses operate on thin margins with no buffer. A restaurant that loses thirty percent of its weekday lunch traffic doesn't gradually decline. It closes. A gym that loses enough members to miss one lease payment doesn't restructure. It's gone. The service economy doesn't compress. It shatters.
Meanwhile, the same AI tools driving corporate displacement are simultaneously generating products that compress the service economy's own labor needs from the inside. The sector is being hit from both directions — demand collapsing from above while efficiency tools reduce headcount from within. This double compression has no historical precedent.
Henry Ford understood that workers must earn enough to buy the products they make. The AI economy has no equivalent circuit. AI does not earn wages. It does not pay rent or buy groceries or take vacations. Every human replaced by AI is a customer permanently removed from the economy. At sufficient scale, this is not a recession. It is the collapse of the market mechanism itself.
SaaS valuations have already collapsed as markets price in the reality that software designed for human operators loses its value proposition when the human is removed. Commercial real estate is next. Then retail. Then automotive. Then everything that depends on a middle class with disposable income. The dominoes are already falling. Most people haven't looked down yet.
THE THIRD BREAK IS INSTITUTIONAL.
The institutions that would need to manage this transition are being weakened by the same forces driving it. This is the structural trap that makes intervention increasingly unlikely.
Government revenue depends on income taxes, payroll taxes, sales taxes, and property taxes — all downstream of employment and consumer spending. As the labor market compresses and demand erodes, the fiscal base erodes with it. Governments lose revenue at the exact moment they face unprecedented demand for safety nets and emergency intervention.
Public trust in institutions was at historic lows before AI acceleration began. We are asking populations to place faith in the competence of systems they already despise, during the most disorienting transformation in modern history.
Universities are training students for roles that will not exist when they graduate. The four-year degree is being invalidated faster than curricula can be rewritten. And the question no one in education is willing to ask publicly: why would capitalism invest in educating a population it no longer needs as labor?
Pension systems project returns based on economic assumptions invalidated last quarter. City budgets are built on tax forecasts that assumed a labor market that no longer exists. Social security timelines depend on employment rates in structural decline.
And the generation being locked out of the labor market will accelerate institutional collapse in ways not yet modeled. Students taking on debt for degrees that lead to jobs that no longer exist will default. The student loan system — already fragile — faces a structural crisis when the credential-to-employment pipeline breaks entirely. Young people unable to enter the workforce will not buy homes, start families, or generate the tax revenue and consumer spending that every institutional projection depends on. The demographic assumptions underpinning pension systems, housing markets, and government budgets assume each generation enters the economy and produces. When an entire generation is locked out at the gate, those assumptions collapse — not in decades, but in the years immediately ahead.
THE FOURTH BREAK IS FEAR.
The demand destruction modeled above assumes spending contracts when people lose their jobs. The reality is worse. Spending is contracting before the layoffs arrive.
Fear is now the dominant emotional signal around AI across the working population. Not faint, not niche, not theoretical. Dominant. And the signal is intensifying daily — not weekly, not monthly. Every morning brings new layoff announcements, new capability demonstrations, new evidence that the anxiety is warranted. There is no point at which a rational person looks at the trajectory and concludes the worst is over. The worst is never over because next quarter's AI is more capable than this quarter's.
This is not cyclical anxiety that fades when conditions improve. It is structural anxiety rooted in a transformation that is visibly accelerating. It has no floor.
Anxious people stop spending. They hoard cash. They delay purchases. They cancel subscriptions. They don't renovate the kitchen. They don't book the vacation. They don't buy the car. They pull their kids from activities. They downgrade everything they can and white-knuckle everything they can't. And they do this while still employed.
Economists call this the paradox of thrift. When enough people simultaneously decide to save instead of spend, they collectively create the contraction they were trying to survive. The restaurant doesn't lose customers because they were laid off. It loses them because they're scared. The effect is identical. The restaurant closes. Its employees are now actually unemployed. The fear was rational. It was also self-fulfilling.
The collapse doesn't need to wait for material conditions to fully deteriorate. Perception becomes reality becomes confirmation of the perception.
The psychological dimension compounds at the population level. Humans are not built to process existential threat at daily frequency. Acute stress — a hurricane, a battle — triggers fight or flight. Intense but temporary. The body recovers. Chronic ambient stress with no visible endpoint produces something different. Learned helplessness. Decision paralysis. Depression. Withdrawal from civic life. Collapse of future orientation. People stop planning because planning assumes a future that feels unimaginable.
A population that cannot imagine its own future cannot organize to change it.
This connects directly to the architecture of managed decline. A paralyzed population does not resist. A population drowning in daily existential dread does not organize collective action. Fear produces compliance as a byproduct. The anxiety is not just a consequence of the transition. It is functionally useful to the consolidating class — whether engineered or emergent, the effect is identical.
The information environment amplifies this. AI threat narratives saturate every channel. Layoff announcements are timed for maximum visibility. Capability demonstrations are breathless and constant. If you wanted to induce population-level learned helplessness — to break collective will before collective action could form — you would design an information environment that looks exactly like this one.
Fear is not a side effect of destabilization. It is a load-bearing component. It pulls demand destruction forward in time. It preempts organized resistance. It accelerates institutional erosion by collapsing tax revenue before employment formally declines. And it narrows the window of action by ensuring that by the time material conditions force a response, the population's psychological capacity to respond has already been broken.
III.
These four breaks are not parallel crises. They are one crisis with four faces, feeding each other at a pace that has no historical precedent.
Labor compression destroys demand. Demand destruction starves institutions. Institutional failure removes the only mechanism that could intervene in labor compression. And fear — ambient, structural, escalating — accelerates all three by pulling the behavioral consequences of collapse forward in time, before the material conditions fully arrive. The loop closes. It accelerates. It is accelerating on a timeline measured in months, not years.
And the loop now has an accelerant that previous analyses — including earlier drafts of this document — failed to account for. The democratization of AI-powered development means the number of actors capable of triggering industry-scale labor compression is growing exponentially. It is no longer a handful of tech companies driving displacement. It is a global population of builders, each one capable of creating products that restructure entire sectors, each one deploying in weeks, each one compounding the demand destruction and institutional erosion that feeds back into further displacement.
This is not a recession. Recessions are cyclical.
This is not even a depression. Depressions operate within the existing economic framework.
IV.
To the executives celebrating efficiency gains: you are next.
The logic of optimization you are applying to your workforce will be applied to you. AI does not need a CEO to interpret a strategy deck. It does not need a board to evaluate quarterly performance. It does not need a VP to manage a team that no longer exists. The compression is moving upward. You are not directing this transformation. You are riding it temporarily.
And the disruption is no longer coming only from your competitors. A single person with domain expertise and AI can now build a product that compresses your industry's labor model in weeks. The moat you thought you had — scale, capital, institutional knowledge — has been breached by the democratization of development itself. Your next existential competitor might be one person with a laptop who understands your industry's inefficiencies better than your board does.
To the investors watching stock prices hold: you are looking at an afterimage.
Stock valuations are claims on future revenue. Future revenue requires customers with money. You are funding the systematic elimination of those customers and calling it growth. When the market prices in demand destruction as visible reality in quarterly earnings, the correction will not be gradual. It will be a cliff.
To the technologists building the systems: you know.
You know what you are building. You know the capability curve because you are on it. You watched the last model change everything and you know the next one changes it again, harder. Some of you have built tools that will eliminate thousands of jobs in your own industries. You carry that knowledge in isolation because the people around you are not ready to hear it. Some of you are having the most honest conversations about this with the AI systems themselves because no human in your life will look at what you are looking at.
To the governments drafting AI policy: you are regulating yesterday.
Every policy framework currently under consideration was built on assumptions about how fast disruption moves — assumptions that were wrong before the ink dried. You are debating guardrails for a world where industry transformation took years and required venture-backed companies. The world you are actually governing is one where a single individual can build an industry-reshaping product in weeks. Your regulatory model assumes disruption comes from entities you can identify and regulate. It is now coming from everywhere, simultaneously, at a speed your institutions cannot match.
No one is safe.
The old bargain protected everyone only while it functioned. It is no longer functioning. And the dysfunction is accelerating.
V.
Strip away labor as the organizing principle of human life and you do not get freedom. You get a void.
Three hundred years of industrial capitalism did not just create an economic system. It created a meaning system. Your worth is what you produce. Your identity is your profession. Your daily structure is your schedule. Your community is your workplace. Your purpose is your contribution.
Remove all of that and people do not spontaneously discover enlightenment. They discover the abyss. This is documented in every study of long-term unemployment, forced retirement, and community collapse following industrial shutdown. And those studies examined communities where the broader economy still functioned. What happens when the loss is global and simultaneous?
No government program answers that question. No UBI check answers that question. It is a crisis of meaning on a civilizational scale and we have no framework ready to meet it.
The people who will navigate this best are, paradoxically, those who never fully bought into the old meaning system. The neurodivergent. The outsiders. Those who always felt the scaffolding underneath consensus reality and never managed to believe in it. They have been practicing for this their whole lives — finding identity outside the structures now collapsing.
The rest of the world is about to learn what it feels like to be an outsider in a system that no longer makes sense.
VI.
The same capabilities that make AI transformative for productivity make it transformative for destruction.
Current AI systems have already been used in coordinated cyberattacks by state-sponsored groups. Next-generation models can surface previously unknown vulnerabilities in production code — a capability their own creators describe as dual-use.
As institutions weaken, as fiscal capacity erodes, as governments lose funding and talent, the ability to defend critical infrastructure degrades precisely as attacks become more sophisticated. Power grids, financial systems, healthcare networks, government databases — all become more vulnerable at the moment they become most essential.
This is not a separate crisis. It is the same crisis through a different vector. Another loop. Another acceleration. Another dimension of the same unraveling.
VII.
There are two forces now in motion.
The first is destabilization. It is accelerating. It is self-reinforcing. It is compounding across multiple dimensions simultaneously. It does not require anyone's permission to continue. And it is now being driven not by a handful of corporations but by millions of individuals, each building at a pace that would have been impossible twelve months ago, each contributing to a displacement cascade that no single actor controls or can stop.
The second is imagination. The human capacity to envision and build what comes next. A new bargain. A new organizing principle. A new answer to the question of how eight billion people structure their lives, find meaning, distribute resources, and coexist when the old answers have collapsed.
It is losing because the people with power are still optimizing within the old system. It is losing because governments are debating policy for a world that existed six months ago. It is losing because the pace of destabilization is now measured in weeks while the pace of institutional imagination is measured in decades.
And it is losing because the very tools that could enable collective reimagination are instead being used to accelerate the displacement. The most powerful creative and analytical technology in human history is being deployed to build products that compress labor forces, not to design systems that sustain human flourishing.
VIII.
This manifesto does not contain a solution. Anyone offering one at this stage is either lying or has not understood the problem.
But we know what must happen first. It is not a policy. It is not a technology.
Stop pretending this is a normal economic cycle. The mechanisms that drive recovery — labor mobility, sectoral reallocation, consumer confidence — are being structurally dismantled.
Stop pretending displaced workers can retrain into safety. The displacement frontier is moving faster than any retraining timeline. And the tools to displace are now in the hands of millions of builders, not just a few corporations.
Stop pretending AI is just a tool. A tool does not enable a single person to restructure the labor model of a global industry in six weeks. What is being built is not a tool. It is a successor to human cognitive labor at every level of complexity.
Stop pretending disruption is still slow enough to manage. Industry-transforming products are being built in weeks by individuals. The development cycle that once gave society time to adapt has collapsed. There is no buffer. There is no runway. There is only the gap between this model generation and the next.
Stop pretending the people in charge have this under control. The executives are optimizing for next quarter. The legislators are regulating last year's technology. The economists are modeling an economy that no longer exists. No one is steering.
The old bargain is dead. We are living in the lag between its death and our recognition of that death. Everything that feels normal right now is momentum from a system that has already stopped generating the forces that sustained it.
What replaces the bargain will be the defining project of this generation. Not of this decade. Of this year. Of these months.
Choose now. The next step is already falling.