by The claim circulating in policy papers, venture capital essays, and parental anxiety threads runs like this: no child born this year will grow up to be smarter than artificial intelligence. The line gets used as a slogan, which is the first sign it deserves examination. Slogans that move easily through dinner parties usually carry hidden machinery. The machinery here is a definition of intelligence narrow enough to fit on a benchmark and broad enough to terrify a parent. Both functions are intentional, and both deserve to be unbundled before the consequences can be argued honestly.
A six-year-old can pour milk without spilling, recognize her grandmother by the sound of her walk on the stairs, and read her father’s mood from a quarter-second facial flicker before he speaks. No current AI does these reliably, which is why the warehouse, the construction site, and the elder-care ward continue to employ humans at rising wages while law firms cut their summer associate classes. What machines do well, with present technology, is symbol manipulation at scale: text, code, formal reasoning, pattern completion across enormous corpora of written human output. The honest version of the claim is narrower than the slogan and still consequential. No child born this year will outperform machines at symbol manipulation, retrieval, or formal reasoning across most of the tasks that currently pay a salary in an office. The slogan compresses that into a panic, which is bad rhetoric and bad policy, and the underlying observation remains true. What follows from the observation is the actual subject of the analysis below.
The Credentialed Class Loses Its Logic
The first casualty is the credentialed professional class, roughly the top 20 percent of American earners by household income. This stratum organized itself across the twentieth century around cognitive screening. The SAT in 1926, refined through the GI Bill expansion. The LSAT in 1948. The MCAT in its modern multiple-choice form in 1962. The USMLE consolidated in 1992. Each gate selected for a particular form of paid cognition: rapid pattern recognition under time pressure, short-term retention of densely structured information, formal reasoning across domain-specific symbol systems. The gates were effective because the cognitive work they screened for was scarce, expensive to develop, and economically valuable.
Three conditions held the system together. Scarcity was the first: only humans could perform the cognitive work, and only some humans, after long training. Expense was the second: the training cost time and money and required institutional infrastructure no individual could replicate. Value was the third: the market rewarded the work because nothing cheaper could produce equivalent output. All three conditions are now eroding simultaneously. A subscription that costs less than a Manhattan dinner produces legal memos, differential diagnoses, and tax planning at a level competent enough to embarrass the junior tier of every paid profession.
Embarrassment falls short of replacement. The senior partner still signs the brief. The attending physician still admits the patient. The accounting principal still files the return. What has collapsed is the economic logic of the apprentice tier, the rung at which young people once learned the trade by performing the work that AI now performs faster and at a thousandth of the cost. Without the apprentice tier, the senior tier has no successors, and the senior tier itself ages out within twenty years. The professions are not being replaced. They are being denied a generation, which is the same outcome on a longer clock.
The lawyer keeps courtroom presence, client relationship, and signature liability. For the doctor, what survives is touch, witness, legal accountability, and judgment under stakes. The architect’s irreducible work happens in the kitchen, in conversation with the homeowner about how the family actually lives. Three of those four functions are not why medical school costs $300,000. The training, the credentialing, the expensive cognitive certification, was effective because it produced the rare commodity. When the commodity is no longer rare, the price of training cannot hold. Either tuition collapses, which would gut the universities that have leveraged themselves on that revenue, or graduates default on debt for credentials that no longer command premium wages. Both outcomes are visible in early data. Neither has yet been admitted by the institutions whose survival depends on denying it.
The same compression is hitting working-class employment, particularly in transportation, customer service, and routine clerical work, and the human stakes there are larger in absolute terms. The reason this analysis concentrates on the credentialed class is that this class produced and sustained the public sphere through which the broader transition will be argued, named, and contested. When that class loses its grip on its own coherence, the conversation about every other displacement becomes harder to organize.
The Parental Project Loses Its Currency
The second consequence is psychological and reaches beyond economics into the structure of family life. American parenting in the educated class has run for at least three generations on a transmission model. Cultivate the child’s mind, secure the child’s place. The cultivation produced status, the status produced security, and the bargain held because each generation could roughly verify the prior one’s judgment. A father who tutored his daughter in algebra in 1995 watched her, twelve years later, take a meeting with someone who had been tutored similarly by similarly anxious parents. The investment paid out in a recognizable currency.
The currency has been redenominated without warning. A father in 2026 watches his daughter receive better tutoring, free, from a machine that has read every algebra textbook ever written and never tires. The democratization is real and worth celebrating. The disappearance of his comparative advantage is also real, and both arrive on the same Tuesday. He had counted on that advantage. Greed had nothing to do with it. The entire architecture of middle-class American parenting had encoded the cognitive premium as the path, and he was a competent parent walking the path his own parents had walked. The consolation that “my child will think for a living” has lost its meaning. What replaces it has not arrived. The vacuum is producing the parental anxiety that fills bookstores, podcast feeds, and pediatric psychiatry waiting rooms, and producing it faster than the helping professions can absorb the demand.
The School System Confronts Its Cover Story
The third consequence runs through the school itself. American schooling has carried at least four functions through the twentieth century: childcare for working parents, social formation, cognitive training, and credentialing for the labor market. The cognitive training and credentialing functions are the two AI most directly displaces, and they happen to be the two schools advertise in their mission statements as the reason for existing. Childcare and social formation remain, untouched and irreplaceable, and no school district raises a tax levy on those grounds.
The honest reckoning is one administrators are not yet willing to give. We run schools mostly to keep parents working and to teach children how to negotiate the social geometry of a room full of other children. The cognitive content has always been somewhat ornamental, a respectable cover story for an institution whose deeper functions were custodial and socializing. AI is forcing the cover story to retire. At least a decade of denial will follow. Curriculum committees will add “AI literacy” units that are structurally indistinguishable from the typing classes of 1985, the computer lab visits of 1995, and the laptop initiatives of 2010, each of which functioned as institutional reassurance rather than pedagogical substance. After the denial, a slow and reluctant rewriting of mission statements will move toward something more honest about what schools actually do, which is gather children safely while their parents earn a living and teach them to sit in rooms with people they did not choose. Both functions are valuable. Neither justifies the per-pupil expenditure of the current system, and the public will eventually discover that the math no longer works.
The Political Bargain Loses Its Foundation
The fourth consequence is political and may be the most important one in the medium term. Technocratic liberal democracy, the regime under which most readers of this essay have lived their entire lives, rested on a quiet bargain. Experts would govern the complicated parts. Voters would govern the simple parts. The experts held position because they knew more than the voters, and the voters tolerated the experts because the system, on average, delivered rising material conditions. The bargain frayed before AI arrived, evident in the populist movements of the past fifteen years, but AI removes the bargain’s foundation outright. If a machine knows more than the expert and the voter alike, the expert has no remaining claim that distinguishes her from any other citizen. She becomes one more citizen with opinions. The voice of trained competence has gone elsewhere, into the model and the dataset, where no human can claim it as her own.
Two political responses follow, and both are visible in the present. The populist response decides that if no human is more qualified than any other, then will, identity, and tribal allegiance settle the question. This is the shape of politics in much of Europe, the Americas, and parts of Asia at the moment of writing, and the authoritarian movements within that response are gaining institutional ground rather than losing it. The technocratic response in a new key hands the decisions to the machine itself, which is the direction parts of finance, military targeting, and judicial sentencing are already moving. The first response sustains the form of democracy while emptying its substance. The second response abandons even the form. Neither response preserves democratic self-rule as the founding generations understood it, and there is no third response visibly forming. The honest political forecast is that what we have called liberal democracy will continue to use its old vocabulary while operating on different machinery, and the gap between the vocabulary and the machinery will widen until the vocabulary collapses, probably within a generation. Whether the collapse opens onto a new democratic form or onto its successor is the open question of the next twenty years.
The Cultural Layer Has Absorbed Shocks Like This Before
The fifth consequence is cultural and harder to predict, because culture has absorbed previous shocks of this kind. Photography arrived in 1839 and was widely expected to end painting. Painting survived by abandoning the territory photography claimed and inventing impressionism, then cubism, then abstraction. Recorded music arrived around 1900 and was expected to end live performance. Live performance survived by becoming an experience economy where presence, not fidelity, was the product. Chess engines surpassed human grandmasters in the late 1990s and were expected to kill the game. Online chess is now larger than at any point in its history, with more humans playing more games against more opponents than the pre-engine era could imagine.
The pattern across these examples is consistent. Mechanical reproduction shifts the value of the human version from product to presence. A handmade chair is no longer a better chair than a factory chair, and it costs ten times more, because the value lives in the maker’s hand and the buyer’s relationship to it. Live theatre does not compete with film on visual spectacle and does not need to, because the live audience pays for the breath in the room. Human writing, if AI writing becomes competent and ubiquitous, will likely become a luxury good signaling effort, time, and personal stake. The author’s life will count for more, and the work without an author behind it will lose value as it becomes plentiful. Whether that economy supports as many writers as the previous one is a separate question, and the answer is no. The professional middle of the writing trade, the working journalist, the staff editor, the workmanlike novelist, will thin out. The top will hold and the amateur base will expand. The middle was always the most vulnerable layer in any cultural economy, and AI accelerates a contraction that began with the collapse of newspaper revenue around 2007.
The Counter-Case Worth Holding
A counter-case deserves to be kept in view, because the foregoing analysis can slide into a fatalism the evidence does not support. Intelligence, as humans have meant the word for most of recorded history, has always carried more than symbol manipulation. The fuller meaning includes desire, mortality, embodiment, the capacity to lose, the capacity to refuse. A chess engine plays better chess than any human and cares about nothing. A writing engine produces fluent prose and risks no humiliation when the prose fails. The child born this year will live in a body that ages, will love people who die, will choose between options under genuine uncertainty about her own future, will know what it is to be afraid without being shut down for it. All of that registers as full-weight human activity, equal in importance to whatever the machine produces. The category is different from symbol manipulation, and the question of which category we will continue to honor with the word intelligence is a political question more than a technical one. The answer will be settled by what the courts protect, what the schools teach, what the markets pay for, and what the surviving institutions of self-government decide to defend.
The Hardest Truth
The hardest truth, the one this site has been documenting across a decade of work on institutional collapse, is that societies do not adjust gracefully to shifts of this size. Institutions built on one logic do not refactor themselves when the logic changes. They hollow out, keep their letterhead, draw their salaries, and lose their function while everyone with standing to name the loss benefits from its concealment. The American university, the credentialing professions, the editorial gatekeepers of the legacy press, the expert commentariat on broadcast television, each is running on borrowed legitimacy at this moment. None of these institutions will announce its own obsolescence. Each will continue to charge tuition, bill hours, issue credentials, and accept underwriting for some years past practical relevance, then collapse when a critical mass of clients notices they have been paying for what is now free.
The collapse will look like the late stages of American public broadcasting documented in the third volume of the Institutional Autopsy trilogy: a long, dignified fade that no one with authority is willing to name in real time, followed by a sudden insolvency event that surprises no one in retrospect. The next fifteen years will involve a generation-long restructuring of who has standing to speak, who deserves to be paid, and what humans are for once the symbol work has been outsourced. Some of that restructuring will be fair. Much of it will be brutal. Almost none of it will be planned, because the institutions in best position to plan are also the institutions with most to lose by acknowledging the situation.
What Is Left for the Child
The children in question will inherit the result without having known the previous arrangement. They will not mourn what they never had. That is the only mercy on offer, and it is offered only to them. The rest of us, who knew the cognitive bargain when it functioned and built our lives on its assumptions, will spend the remainder of our working lives attending its funeral while pretending it is still in business. The pretense will be socially mandatory, professionally protective, and personally corrosive.
The honest response is to name what is happening, refuse the pretense, and locate value where it is actually moving, which is into presence, judgment, embodiment, and the kind of human authorship that machines cannot fake because they have no stake in the result. The child born this year, if she is lucky, will grow up in a world that has finished the funeral and started building the next thing. The question is whether her parents and grandparents can endure the funeral with enough dignity to leave her something to build on.
Leave a Reply