Artificial Intelligence in the Longue Durée
The dawn of artificial intelligence has unleashed equal measures of optimism and fear regarding its potential to fundamentally transform society. As AI becomes increasingly widespread, a crucial question presents itself: Is this the beginning of a new arc of change, or merely an extension of an existing trajectory?
Consider the life of Thomas Avery. He was born in 1865 in rural Burke County, Georgia, to a poor white farming family. His early childhood unfolded in the immediate aftermath of the Civil War and during Reconstruction, in a community where the legacies of slavery remained firmly entrenched. White mobs attacked freedmen. His father’s plow was driven by mules, his mother cooked over an open fire, and although the cash economy had taken hold, much local economic activity still took place by barter. He attended school for just two years before joining his father and older brothers in the cotton fields.
As a young man, Thomas left the South during the first wave of internal migration and settled in Chicago, where he found work in a Union Stockyards meatpacking plant and later in a steel mill. He married and raised four children in a crowded South Side tenement. In his 50s, he transitioned into clerical work at a growing insurance company. There, he learned to type, managed office records, and eventually supervised a small team of administrative staff. Over time, he adjusted to a life marked by indoor plumbing, automobiles, wage pay, electric lighting, organized labor, and the emergence of a basic social safety net.
By the time Thomas died in 1945 at age 80, he had not only seen electrified homes, airplanes in flight, and the birth of mass media, but also the Great Depression, two world wars, the Holocaust, and the atomic bomb. He was born into a life that a 15th-century peasant would have largely recognized; he died in one broadly familiar to a modern person.
Now consider Mary Gonzalez, born in 1945 in Newark, New Jersey, to a Puerto Rican family that arrived on the mainland during World War II. Her father worked as a high-school janitor while her mother cleaned suburban houses. Mary graduated from high school in 1963 and began working as a legal secretary in Manhattan after earning a certificate at Union County College. She married in 1968, bought a modest split-level home on Long Island, and had two children. In her 30s, Mary returned to school at night and completed a bachelor’s degree in business administration. She pursued additional coursework at the graduate level, taking night classes and then online classes in employment law and organizational behavior. Over the course of her career, she rose through the ranks at a mid-sized law firm, becoming a senior paralegal and later the firm’s human-resources head.
When she died in early 2025 at age 80, Mary had seen quite a bit of change. She lived through the civil-rights movement, the Cold War, the Vietnam War, the expansion of suburban America, the women’s-liberation movement, the 9/11 attacks, the rise of the internet, and the dawn of AI. But she also spent her entire life in a world with public high schools, automobiles, commercial air service, big corporations, a stock market, near-universal wage labor, mass media, the same political institutions, and broadly familiar legal, cultural, and commercial norms.
Comparing these two fictional lives highlights a crucial distinction: Though each spanned eras of great change, the 1865-1945 period that Thomas lived through reshaped the deeper structures of economy, culture, and society in ways that more recent decades, for all their upheaval, have not.
There’s a chance we’re on the verge of another transformation. AI is no futuristic abstraction; it is already a central subject of public policy at every level of government. Understanding the kind of change it will bring — whether it will reflect Thomas’s era of revolutionary upheaval or Mary’s continuation of an existing evolutionary arc — will be essential to making policy that works.
THINKING ABOUT HISTORICAL CHANGE
For most people in history, changes in ideas, institutions, and technologies — including those emphasized in school curricula as “revolutions” — unfolded slowly, leaving daily life broadly familiar to those who were born and died during those eras. A German farmer born in 1500 might have lived through the Reformation and the spread of printing without leaving his village or altering his daily routines. An English day laborer born in 1760 might have died in 1820 having heard of steam engines and spinning frames while still digging ditches and hauling loads as he always had. These periods mattered, but for most people they looked less like the radical break of Thomas Avery’s life and more like the gradual evolution Mary Gonzalez experienced. These revolutions unfolded unevenly, and often left the basic shape of daily life from birth to death intact.
Such slow-moving transformation contrasts starkly with one of the most discussed ideas in contemporary AI discourse: the singularity. The term dates to the 1950s, when mathematician Stanisław Ulam attributed it to John von Neumann, who reportedly spoke of
he ever-accelerating progress of technology and changes in the mode of human life, which gives the appearance of approaching some essential singularity in the history of the race beyond which human affairs, as we know them, could not continue.
Ray Kurzweil popularized the idea in the 1990s, imagining a future where computing advances so rapidly that humans and machines merge, minds are digitized, and institutions become obsolete. In such a world, policymaking loses its footing entirely. Planning for such a future would be like revising the tax code during a zombie apocalypse. The change would be so sweeping that there would be little policymakers could do to prepare.
But what if AI is unlike the singularity, and more like the age of exploration, the First Industrial Revolution, or the digital revolution that Mary experienced? In that case, the script already exists: Smart, limited regulation; strong research universities; market reliance; and the rule of law may be enough. Though adapting these elements of society to new realities will not be easy, and will require serious thought regarding all sorts of novel questions, the terrain will be familiar. While the world of 2225 might prove unrecognizable to someone today, the changes experienced across a single lifetime would be manageable under the existing systems.
Among the possible trajectories for AI, however, another one deserves particular focus: a structural transformation in the longue durée — the long arc of history — visible within the span of a single human life. This is simply one of many scenarios, but it would demand the most intellectual attention. Unlike a singularity, which defies preparation, and the important historical changes of more recent years, which were managed with familiar tools, a structural shift demands fresh thinking and institutional adaptation. If such a transformation is underway, it will challenge the premises of existing systems and sweep them away with great speed.
Public policy, in this scenario, is not peripheral; it’s central. Structural shifts require not only technical governance, but historical awareness — including an appreciation of how changes in the longue durée reshape how societies work, think, and govern. The sections that follow examine tools for recognizing such transformation, apply them to AI, describe the major cultural trends that would likely signify a shift, and, finally, offer broad policymaking principles.
WHAT IS A STRUCTURAL TRANSFORMATION?
Historians and political theorists have long tried to distinguish between momentary disruptions and typically gradual transformations that ultimately reshape society. The latter is common: Nearly everyone born in the West since a little after 1400 has lived through a period some historians credibly label a historical “revolution” of one type or another. Many of history’s most significant changes transpired over generations through the evolution of existing institutions.
One widely cited — but often misunderstood — framework for explaining transformative change comes from Thomas Kuhn, the mid-20th-century physicist and philosopher who introduced the concept of “paradigm shifts” in scientific understanding. In Kuhn’s view, science does not advance linearly through steady accumulation, but through alternating periods of “normal science” and revolutionary crisis, when mounting anomalies subvert the prevailing paradigm and a new one takes its place. Although Kuhn wrote about scientific changes, his vocabulary has been broadly applied in political and historical contexts — often in unhelpful ways. Social systems, unlike scientific fields, tend to be more diffuse: Their paradigms are less coherent, their anomalies harder to quantify, their transitions rarely abrupt and almost never truly total.
To better grasp the kinds of change that unfold across entire societies, we would be better off turning from the laboratory to the study of history, where figures like Antonio Gramsci and Fernand Braudel developed tools for recognizing transformations as they occur in living institutions.
Gramsci, a communist, offered a genuinely useful analytic distinction. He believed social transformations were typically slow and difficult to perceive but could accelerate in moments of crisis — when old systems lost legitimacy and new ones had not yet solidified. He saw his own time as an instance of what he called an interregnum, when what he saw as the failures of democracy and capitalism rendered the existing order illegitimate. In such moments, he argued, new hegemonies could emerge. Though his hoped-for communist revolution never materialized, his understanding of crisis and structural realignment remains a useful tool.
Braudel, on the other hand, distinguished among three levels of historical time: short-term events, medium-term cycles, and long-term (la longue durée) structural forces, emphasizing that real transformation often unfolds in the rhythms and patterns of work, rest, trade, family, and migration. The longue durée involves changes to the deep structures of everyday life that typically evolve slowly, reshaping the boundaries of what feels permanent. Braudel, though often associated with slow-moving transformations, also acknowledged that in rare but decisive moments — such as wars, revolutions, or climatic disruptions — forces within the longue durée could move quickly, revealing just how fragile long-standing routines might be.
His account differed from Gramsci’s in important ways. Whereas Gramsci emphasized ideological and institutional rupture, Braudel focused on embedded routines and material practices. But while their emphases differed, both thinkers offered frameworks for understanding how crises can accelerate latent structural change.
Others have extended or paralleled these insights. Michel Foucault, the French historian and philosopher, described what he called epistemes — the foundational, often unconscious structures of knowledge and value that shape what a society permits itself to see — and argued that, thanks to these phenomena, only certain forms of knowledge, authority, and organization are thinkable in a given era. Meanwhile Karl Polanyi, a 20th-century Hungarian economic historian and social theorist, argued that major economic shifts trigger a “double movement”: the expansion of markets, followed by a counter-response seeking to restore social stability. These dynamics, he suggested, often reshape institutions. His insights complement the work of Gramsci and Braudel by emphasizing that structural transformation rarely unfolds in a single direction.
Though vastly different in emphasis, all of these thinkers grappled with the same question: How does change work at the deepest level, and what are its signatures? Whereas Gramsci analyzed ideological legitimacy and Foucault interrogated epistemes, Braudel listened for change in the patterns of daily life, and Polanyi tracked institutional responses to economic upheaval.
These lenses are not crystal balls, but they offer ways of detecting institutional realignments. They help clarify the difference between sweeping transformations, like the one Thomas Avery experienced, and more gradual shifts, like those that occurred during Mary Gonzalez’s life. They suggest that structural shifts often begin at the margins — through accumulating anomalies in norms and expectations, as well as institutional friction — signaling that deeper transformations are in progress.
USING THE FRAMEWORKS
If AI marks the beginning of a structural transformation, the signs are scattered and subtle: collapsing legitimacy; reconfigured routines; and shifts in how work, value, and time are organized. Despite their differences, both Gramsci and Braudel would likely recognize the current moment as one of meaningful disturbance where structural change is possible.
Gramsci’s concept of interregnum maps well onto the past few years. In the wake of the George Floyd protests, elite institutions across the country embraced ideas drawn directly from Gramsci’s own theory of cultural power: the notion that social and political control is maintained not only through coercion, but through the institutions — schools, media, civil society — that create cultural consent.
These ideas — often described under the rubric of systemic oppression or ideological reproduction — did not emerge in isolation; they were developed and extended by critical race theorists such as Kimberlé Crenshaw, who introduced intersectionality, and Angela Davis, who argued that structures of race, class, and gender are interwoven through the institutions of American life. The rapid adoption of their frameworks — by universities, non-profits, Hollywood, media outlets, and large corporations — operationalized Gramscian cultural analysis on a mass scale.
But that embrace generated a backlash. Parents organized to reject new educational materials. State legislatures passed laws targeting diversity, equity, and inclusion programs. Donald Trump’s 2024 campaign triumphed by explicitly attacking this ideology, portraying it as elite overreach. What followed was a back-and-forth: institutions swinging from affirmation to repudiation.
Herein lies the irony: That there exists a fierce fight over Gramsci’s contentions about systemic oppression says nothing about the validity of those ideas, but a great deal about his theories of structural change. Gramsci argued that cultural consensus is never fully settled — that legitimacy is always under contest, especially during interregna. The volatility of institutional commitments — expressed as the oscillation between adoption and disavowal — offers a near-perfect example of the instability he sought to describe.
AI has become both a terrain for this ideological combat and a force that accelerates it. From early controversies over image generators that refused to depict certain historical figures as white men, to critiques of alleged algorithmic bias in criminal sentencing or hiring software, AI has been pulled into broader debates about identity, fairness, and cultural norms. A 2024 Pew Research Center survey found that while over 75% of AI researchers viewed the technology with optimism, less than 25% of the general public shared that view. This isn’t simply a knowledge gap: It reflects a deeper dispute over the values, power, and trajectory embedded in the technology itself.
AI is appearing — often chaotically — within the very institutions Gramsci highlighted. A 2024 RAND study found that only 18% of U.S. school districts reported offering formal guidance on AI use during the 2023-2024 academic year, but another study suggested that 84% of teachers already used AI tools in some way. Journalists, too, are experimenting with AI authorship. At the Washington Post, the Ripple initiative deploys an AI writing assistant, Ember, to help readers submit opinion pieces.
Even the state itself is in the mix. The “Make Our Children Healthy Again” report, which attracted scrutiny for citing fictitious AI-generated sources, illustrated how rapidly institutions can come to rely on powerful technologies they barely understand. These developments may erode institutional legitimacy and open up space for counter-narratives — what Gramsci would call the fraying fabric of consent.
Braudel, by contrast, asks us to look elsewhere — not at ideological skirmishes, but at the long drift of habits, routines, and built environments. He studied the persistent structures of everyday life that shape how people work, move, trade, and experience time. Here, too, we’re beginning to see major shifts, many driven directly by AI.
We can begin with the realm of work. The Covid-19 pandemic, combined with advances in telecommunications technology, restructured where and when white-collar work happens. That shift has not reversed: During the first quarter of 2024, 23% of U.S. workers performed some or all of their work remotely. Meanwhile, 70% of businesses now invest in AI-powered collaboration tools. These are not just efficiency upgrades; they remove geographic boundaries and alter the rhythms of the workday.
At the same time, AI is changing what work is. Tools like Cursor and Cody for engineers, CoCounsel for lawyers, Firefly for designers, and Glass AI for doctors automate knowledge work once done by junior staff. While this has prompted some concern about the “disappearing ladder” — worry that entry-level roles are being rendered obsolete — we’ve clearly not arrived at that future yet. Unemployment remains low by historical standards, and while some career pathways are shifting, AI has, so far, created more jobs than it has eliminated. We are seeing real change in the structure of work, but there’s little evidence of its elimination or any radical transformation.
Consumption is shifting, too. Some of this transition is occurring along class lines: High-income households increasingly rely on AI shopping assistants, personalized delivery platforms, and subscription-based services across health, education, and entertainment. These technologies allow for hyper-curated lifestyles and further embed AI in everyday decision-making.
Lower-income households, on the other hand, are increasingly turning to Buy Now, Pay Later (BNPL) services to manage essential expenses. A 2025 LendingTree survey found that 25% of BNPL users financed groceries using these services, up from 14% the previous year. Bank of America data indicate that nearly half of BNPL users in March 2024 earned less than $50,000 annually. These services, which provide credit to people banks would not consider creditworthy, are themselves driven by sophisticated, individualized AI models.
While trends differ according to class, people are buying more digital goods and services across income levels. AI is a big part of this story. Consumers now subscribe not just to entertainment platforms, but to AI-enhanced tools such as language tutors, fitness apps, writing co-pilots, and virtual stylists.
These behaviors suggest a reorganization of consumption. Braudel would note that the way people acquire, value, and exchange goods is changing — not just through price or preference, but through a restructuring of what it means to consume in a mediated world.
Time is also compressing, thanks largely to AI. Braudel understood structural transformation in part as a reordering of time — how people allocate it, experience it, and coordinate it socially. Today, AI tools promise efficiency gains. A 2024 Thomson Reuters report found that legal and financial professionals expect to save up to 200 hours per year through AI automation. But these time savings can mask new obligations. The same tools that streamline tasks can also expand expectations — shrinking deadlines, encouraging 24/7 availability, and making downtime feel unproductive.
Braudel would likely see in this a transformation of temporal experience itself. As tasks speed up and expectations multiply, the rhythms of everyday life are diffused. These trends are not new, but AI will likely accelerate them.
Gramsci and Braudel rarely move in sync: One waits for the crisis of ideology; the other watches for reorganization in life’s scaffolding. But when both register signs of strain — when legitimacy breaks down and daily routines are quietly rearranged — the convergence is significant. It doesn’t guarantee that we are on the verge of a structural transformation, however. Braudel might note that though many patterns suggest structural change, much of the longue durée still appears intact. He would be quite intrigued, but not yet persuaded, that long-term structural change is taking place. Gramsci, by contrast, would see in the cultural fragmentation and institutional instability clear signs of a new hegemony being born.
Thus, our analysis using Braudel and Gramsci’s models is instructive, but inconclusive. Since change may arrive in many ways, it is worth drawing on historical precedent — particularly periods of rapid change in the longue durée — to help fill out the picture of structural transformation sketched so far.
LESSONS FROM HISTORY
Structural transformations do not merely introduce new tools or industries; they reverse social truths, reorder time, and redefine the state’s fundamental role. Among the many possible examples, two stand out as the clearest and best-documented cases of deep structural change within a single human lifetime: the aftermath of the Black Death in 14th-century England and most of Western Europe, and the Second Industrial Revolution of the late 19th and early 20th centuries. They illuminate a kind of transformation beyond mere disruption, reaching into the deepest rhythms of social, economic, and political life and fundamentally changing the structure of the typical person’s day-to-day experience.
The Black Death shattered a core assumption of medieval European society: that labor was immobile, hereditary, and perpetually tied to land — the major source of wealth. For centuries, serfdom had underwritten the social order. Armies, courts, churches, tax systems, and markets all relied on the idea that the vast majority of people would remain bound to their place and station. Serfs could not leave the land as a matter of law; peasants rarely did because few real opportunities existed.
When the plague wiped out 30% to 50% of England’s population, this foundation collapsed. The ruling class attempted to enforce the old world through laws like the Statute of Laborers of 1351, which sought to freeze wages in nearly all professions and imprison anyone who tried to switch jobs. But enforcement failed. Eventually, wages rose, migration surged, and feudalism crumbled in England — not by revolution, but by the extinction of the conditions that had once supported serfdom.
Meanwhile, new technologies like windmills, padded horse collars, and improved looms — combined with a growing wool trade — began to create wealth in ways other than land ownership. Serfdom continued as a matter of law in many places but was crippled in England and Western Europe. Truths that had once defined the logic of society — that labor must be fixed and servile, and that capital consists mainly of land — were reversed.
Foucault might interpret these as ruptures in the episteme — moments when the organizing structures of knowledge and governance are fundamentally redefined. The Black Death didn’t just upend feudal labor arrangements; it rendered imaginable the very idea of labor as mobile, transactional, and personal.
The Second Industrial Revolution, centuries later, reversed a more subtle foundational truth, this time about geography: Where people could live became constrained, while capital was liberated from location constraints. Before the Second Industrial Revolution, energy was immobile and deeply tied to place. Large-scale enterprises had to be located near specific sources of power — fast-moving rivers for water wheels, coal fields for steam engines and steel mills. Efficient distribution of goods required access to rail lines or navigable rivers, which meant major commercial operations clustered at just a handful of rail hubs and riverfronts. Steel mills sprang up next to coal fields and iron-ore deposits. In this world, geography dictated economic possibility: Commerce occurred where energy and transport converged.
The internal combustion engine and electrification shattered these constraints. Energy could now be generated in one place and used in another. Trucks could move goods without railroads. Industry siting was no longer beholden to natural features, but to flexible factors like land prices, labor supply, and proximity to other firms. The only real requirement was access to the grid — a minimal threshold compared to the cost of opening a new mine or factory.
Yet as capital gained mobility, individuals became deeply tethered to infrastructure. In earlier eras, farmers and artisans could settle wherever there was water and build viable lives using local materials and basic tools. Towns emerged around wells, crossroads, or train stops, whenever necessity or opportunity allowed. But the modern built environment — shaped by electricity, public sewage, and streetcar networks — redefined viability. Participation in full civic and economic life now required residence within systems built by large entities. Homes without wires, modern plumbing, or telephones were soon considered uninhabitable. The sheer size of enterprises made it impractical for everyone to live close to their place of employment: As many began riding trolleys to work, proximity to a streetcar line also became a major factor.
Here, too, a reversal in the episteme occurred. A society in which capital was essentially placeless and infrastructure was destiny — where firms moved freely while people remained attached to the grid — became imaginable. It had once seemed natural that individuals lived wherever they could build, but this was now a luxury reserved for the rich. Power, in the Foucauldian sense, migrated into infrastructure itself. The electrified grid, the sewer system, and the trolley line were not just utilities; they acted as instruments that altered the fabric of daily life. The Second Industrial Revolution didn’t merely change the built environment; it inverted its logic. What was once fixed became mobile; what was once mobile became fixed.
In addition to remaking the economy, these material reversals reshaped the structure of time itself. As Braudel argued, true historical transformations are often visible not in the drama of politics or war, but in the slow churn of daily life — including how people tell time. While new technologies will alter time in a few lines of work at any given point, the typical person’s daily schedule remains relatively stable from year to year. The Black Death and the Second Industrial Revolution, however, each reordered these deep temporal structures for the median person.
In medieval England, time was governed almost entirely by the rhythms of nature and the Church. While mechanical clocks began to appear in the 13th century — mainly in monasteries and cathedral towers — most ordinary people had never seen one, and few would have any need to know the hour, let alone the minute. Days began with the sun and ended with dusk. Agricultural labor followed the rhythm of the seasons, the weather, and the crop cycle. The dominant form of temporal organization was liturgical and feudal, not commercial: The year was shaped by saints’ days, harvest festivals, fasting seasons, and obligations to one’s liege lord.
Daily discretion over one’s schedule was almost non-existent. Professions were often inherited by law or custom. Social mobility was largely limited to narrow paths through the Church or the military, and even there, one’s family status mattered immensely. Outside the small merchant class and a fringe population of itinerants, mercenaries, and entertainers, nearly all people — including the upper class — lived in tightly ordered, place-bound hierarchies. The rhythms of life were imposed from above by nature and tradition.
The Black Death destabilized this daily schedule’s order as much as it did the labor market. As wage labor expanded, time itself became a commodity that individuals could sell — and increasingly had to. The idea that one might have some say in how to spend the hours of a day became thinkable, then normal. For many, manorial obligations gave way to market incentive: People began organizing their lives around opportunity rather than obligation and taking control of their own leisure time. It was this shift in the daily schedule that had the broadest impact.
Centuries later, the Second Industrial Revolution transformed daily life even more radically — by mechanizing time itself. Before electrification, most everyone had seen a clock, but few needed to know the exact minute. Until railroads introduced standardized time zones in 1883, neighboring towns could operate on different times, set by a church bell or courthouse clock. Work followed the natural world: sunlight, weather, season, water flow, and task at hand. While factories, communal farming, and slavery imposed tighter controls on some, most people were either self-employed or worked in small, flexible enterprises where daily rhythms remained loosely structured.
Communication, too, was asynchronous. Outside of military or telegraph-linked industries, sending a message meant waiting hours, days, or months for a reply. The Second Industrial Revolution changed this dramatically. Electrification, synchronized clocks, and mechanized factories brought with them minute-by-minute scheduling. Factory work, once limited to places in a few big cities, became normal and required precise coordination. New technologies like the telephone and radio enabled synchronous communication across vast distances. The school day was patterned on factory time, complete with bells, rigid periods, and punctual attendance. Cities lit by electric lights no longer slept at sundown. The rhythm of life detached from nature and aligned instead with infrastructure and machines.
Though the era brought major gains in leisure — particularly with the emergence of the two-day weekend — discretion over one’s work time declined sharply. With capital equipment fixed in place, telecommunication still expensive, and coordination essential, working outside the home — and in turn, the commute — became nearly universal. Time, once communal and fluid, became segmented and external. The experience of daily life, once bound to fields and festivals, was now governed by commutes, machinery, grids, and labor contracts. Autonomy at work contracted, even as free time expanded overall.
These transitions — first after the Black Death and again after electrification — did not just introduce new tools or institutions; they redefined what it meant to experience a day, to make a choice, and to plan a life. In Braudel’s terms, they reshaped the everyday infrastructure of existence — that slow, nearly invisible layer of historical change that ultimately proves most enduring.
When truths are reversed and time is reordered, the role of government does not remain the same. In both episodes, the purpose of the state shifted in response to new demands. After the Black Death, elite efforts to restore the old order faltered as enforcement of the Statute of Laborers failed and political legitimacy began to migrate away from land ownership, inheritance, and capacity for violence and toward representation and the commons. The collapse of feudal labor arrangements did not lead directly to democracy, but it eroded the ideological scaffolding of aristocratic rule and set in motion its centuries-long demise.
In the industrial age, this shift was even more explicit. As Karl Polanyi argued, disconnecting markets from traditional social systems provoked a “double movement.” While one force sought to liberalize and expand markets, a counterforce demanded new protections and social anchors. The state became that anchor. It was called on to regulate industry, ensure education, provide health care, manage pensions, and intervene in labor disputes. As families became less able to rely on kinship networks or local custom, public systems were expected to take their place. The state was no longer just an enforcer of contracts or a guardian of borders, but the manager of mass welfare and national infrastructure. This was not just a functional adjustment; it was a redefinition of purpose driven by the double action of disruption and response.
NO STRUCTURAL SHIFT…YET
AI has already begun reshaping economy, culture, and governance. But if we take seriously the tools of Braudel, Gramsci, Polanyi, and Foucault in light of the lessons of history, we can see that AI has not yet triggered a structural transformation. While many clues might indicate society is on the verge of a transformation, little indicates we are already in one. To date, no foundational social truths have been overturned. The rhythms of daily life for most people remain intact. The functions of the state remain largely unchanged.
Yet AI’s potential to reverse long-held assumptions about labor, intelligence, and authority is significant. If AI begins to undermine the idea that human judgment is central to work, that intelligence is inherently biological, or that institutions need fixed professional roles, it may catalyze change as deep as that triggered by the plague or electrification.
At this early stage, the most reasonable policy posture toward AI is humility. It remains unclear whether AI represents a structural transformation on the scale of our historical examples, or merely a powerful — but not unprecedented — continuation of the digital revolution. This absence of certainty is not a weakness of analysis, but a recognition of how deep change often unfolds. The most damaging outcomes will likely come from premature confidence rather than cautious observation.
This humility extends to regulatory design. It doesn’t mean inaction, but rather observation, pluralism in experimentation, and patience in lawmaking. The moratorium on state-level AI regulation that the House of Representatives passed in the spring of 2025 before it was rejected in the Senate reflects sound and safe policy: A fragmented, state-by-state approach to AI rulemaking would burden domestic firms with conflicting obligations, increase compliance costs, and create strategic opportunities for global competitors, especially China.
Recent history supports this view. In the 1990s, the federal government chose not to aggressively regulate the early internet. That decision gave U.S. companies the space to experiment, scale, and set global standards. In the mid-1990s, the American and Eurozone economies were roughly equal in size. Today, the American economy is nearly twice as large. The causes of that divergence are complex, but permissive digital policy played a meaningful role. If AI is a continuation of the same technological arc, a similar regulatory pause could yield disproportionate benefits.
The case for restraint is not purely economic: Regulating too soon will almost certainly mean regulating the wrong things. Structural shifts rarely present their most important effects directly, instead appearing in third- and fourth-degree consequences — in secondary institutions, behavioral shifts, and conceptual misalignments. First instincts about what to regulate often miss the point, even for non-structural shifts.
Consider the 2021 congressional hearing titled “Algorithms and Amplification: How Social Media Platforms’ Design Choices Shape Our Discourse and Our Minds.” Every word in that title would appear in a standard dictionary from 1950. And yet the topics discussed and the reasons why they are socially important would be very difficult to explain even to a well-informed person in 1995. The challenge is not that new technologies are incomprehensible; it’s that their effects outpace inherited frameworks.
If the transformation underway becomes structural, the most relevant policy analogies may lie not in recent statutes, but in deeper history. The Telecommunications Act of 1996 provides a useful model for managing an incremental, non-structural shift, but it may offer little guidance for a transformation that reorders labor, cognition, or institutional legitimacy. In that case, a more instructive lesson may come from the failure of the Statute of Laborers, which tried to impose stability on a world already in motion at an incredible rate. Its failure delegitimized a whole theory of government.
We do not yet know whether AI marks the beginning of a transformation like the one Thomas Avery lived through, or whether it will simply carry forward the slower arc Mary Gonzalez experienced. But that is the central question: not whether AI is powerful or disruptive in some areas, but whether it is quietly rearranging the architecture of daily life, redefining what it means to work, learn, be governed, and know. If a tectonic shift comes, it may be dressed in routine: hidden in software updates, workflow tweaks, and cultural whiplash over important issues. It will not ask for permission.
Still, we must be open to the possibility that many people walking around today will live the life of Thomas Avery, crossing a threshold so consequential that what once felt permanent will become quaint and what once seemed impossible will become mundane.