When Humans Are No Longer Needed
For most of human history, the powerful needed the rest of us. Pharaohs needed peasants to grow the grain. Feudal lords needed serfs to work the land and sons to fight their wars. Industrial capitalists needed workers to run the factories and consumers with wages to buy what the factories produced. Even the cruelest regimes of the twentieth century were constrained by this mutual dependence: Stalin had to stop purging the army because he needed it to function, and the Nazis’ extermination campaign contradicted their own war effort because they still needed labor.
This mutual need was never a moral achievement. It was a structural fact, and it was the floor under how badly elites could treat the rest of us. Unions worked because strikes hurt. Democracy expanded because mass armies required mass buy-in — you cannot draft a million men into trenches without eventually conceding them the vote. The welfare state emerged partly because healthy, educated workers were more productive, and partly because unhealthy, angry workers were a revolutionary threat. The twentieth-century social contract, for all its ugliness and exclusions, rested on a simple material reality: the people at the top needed the people below them to produce, to fight, and to buy.
That floor is being removed.
If artificial intelligence and robotics deliver on even half of what their builders claim, we are approaching the first moment in human history when a small class of people will not need the rest of humanity for anything. Autonomous systems will fight the wars. Automated factories will make the products. Algorithmic targeting will reach whatever consumers remain. The billionaire with robot labor, AI advisors, private infrastructure, and algorithmic content doesn’t need to cooperate with anyone outside his immediate circle, and the arguments that used to pretend he did — that we’re all in this together, that his prosperity depends on ours — were leaning on material facts that are evaporating.
This is the rupture underneath the surface chaos of the current moment, and it deserves to be named plainly, because almost no one in our public conversation is willing to name it.
Think about what we are actually living through. The post-1945 world order — the UN, NATO, the EU, the nuclear taboo, the idea that aggressive war is illegal — was built by people whose formative experience was watching the nation-state kill a hundred million humans in thirty years. Those institutions are now being dismantled in real time, not because they failed but because the generation that remembered why they were built is dying, and their grandchildren find the constraints annoying without understanding the alternative. Freedom House has tracked eighteen consecutive years of net democratic decline. Borders have been changed by force in Europe and the international response did not stop it. The Security Council is paralyzed, the WTO is broken, the ICC’s warrants are ignored by the powers that matter.
At the same time, wealth has concentrated at a pace and scale without modern precedent. The top tenth of one percent of Americans now owns more than the bottom eighty-five percent. Billionaire wealth has roughly doubled since 2020. A famous study by the political scientists Gilens and Page found that in the United States, policy outcomes correlate strongly with elite preferences and essentially not at all with the preferences of ordinary citizens once you control for what elites want. The forms of democracy persist. The substance has been hollowed. This is not a conspiracy theory. It is a finding in mainstream political science that has been largely ignored because the implications are too uncomfortable to sit with.
Layer on top of this the specific conflicts that dominate the news: Israel-Gaza, Israel-Iran, Russia-Ukraine, the Sahel coups, the slow-motion disasters in Sudan and Yemen and Myanmar. These look religious or civilizational because the actors involved speak religious and civilizational languages, and because our media covers them through those frames. But most of them are arguments about borders drawn by Europeans over brandy in the early twentieth century, about resources controlled by foreign firms, about populations trapped inside states they never consented to join. The Kurds are the largest stateless people on Earth because Sykes-Picot split them across four countries in 1916. Iraq is three Ottoman provinces stapled together by the British to control oil. Nigeria combined two hundred fifty ethnic groups, including former rival empires, into a single colonial administrative convenience. The wars we’re watching are the long-delayed consequences of decisions made by men who never visited the places they partitioned, and the religious packaging is what political entrepreneurs use to convert specific grievances into mass mobilization, the way Milosevic converted ordinary Yugoslavs who lived in intermarried apartment buildings into Serbs, Croats, and Bosniaks willing to kill each other.
These things are connected. The same loss of institutional legitimacy that allows borders to be changed by force also allows billionaire wealth to go untaxed. The same information chaos that makes religious weaponization effective also makes antitrust enforcement impossible. The same material pressures that produce wars of identity — climate stress, demographic shifts, resource competition, the hollowing of the middle class — are also the pressures that automation and AI are poised to make dramatically worse. And the cultural flattening that Allan Bloom diagnosed in The Closing of the American Mind almost forty years ago — the production of hollow, rootless, distractible individuals incapable of serious thought — has continued and accelerated in ways Bloom could not have imagined, which means the population facing this unprecedented situation is less equipped to think about it clearly than any recent generation.
It is scary, but here is what I believe we are not saying out loud because it is too frightening to say: the mutual dependence that historically forced cooperation across class lines is dissolving in both directions. The rich no longer need the poor. The poor increasingly don’t need each other. You can order groceries without knowing anyone, work remotely without colleagues, entertain yourself without a community, date through algorithms, get therapy from chatbots, educate your children through screens. Every one of these looks like an expansion of individual freedom, and in narrow senses it is. The aggregate effect is that the web of mutual need that made cooperation non-optional — the factory, the neighborhood, the parish, the union hall, the bowling league, the draft — has been dissolving for forty years and is now dissolving faster. Robert Putnam documented the beginning of this in Bowling Alone a quarter-century ago, and the trends he identified have all accelerated. The share of Americans who say most people can be trusted has fallen from about half in the 1970s to about a quarter today. Trust, once lost, is extraordinarily hard to rebuild, because rebuilding it requires the very cooperation that low trust prevents.
The moral and political vocabulary we’re using to address current problems — solidarity, community, citizenship, public good — was built for conditions that no longer obtain. It presupposed a web of mutual dependence that made these concepts pragmatically necessary as well as morally attractive. If that web dissolves, the concepts don’t automatically dissolve with it, but they become harder to motivate. Why should a wealthy American care about a poor American if their lives no longer intersect in any meaningful way? Why should a young person invest in political institutions that seem not to affect her life? The old answer was that you had no choice. The new answer has to be that cooperation is a better way to live than its alternatives. That is a much harder argument to win, because it requires people to choose cooperation for reasons other than survival, and humans are not reliably good at that kind of choice.
None of this has to end the way it looks like it might end. The scenario in which a few thousand people control automated systems and manage a surplus population through surveillance, welfare, entertainment, and pharmaceuticals — the scenario toward which the current trajectory points — is not historically stable. Every previous attempt at that kind of arrangement has broken down within a few generations, because the internal contradictions are severe. Elite factions turn on each other. Middle strata demand a share. Populations eventually refuse to cooperate with arrangements that treat them as disposable. The oligarchic scenario is a failure mode, not a steady state. But the transitions out of it have historically been catastrophic, and the scale of the technologies now available means that the next such transition could be worse than any in modern memory.
What might prevent this, if anything can, is not a grand political program but something more difficult: the reconstruction of chosen cooperation in a world where structural necessity no longer forces it. Chosen cooperation is what holds monastic communities together in periods of civilizational collapse. It is what built the abolitionist networks, the civil rights movement, the mutual aid societies, the labor organizing in its idealistic phases. It is rarer than necessity-driven cooperation and harder to sustain, but it can be surprisingly durable when it forms around genuine values. It selects for people who actually want to cooperate, rather than people who have no choice, and it can preserve what matters through periods when the larger structures fail.
The practical form of this, at the level where ordinary people can actually act, looks unglamorous. It is small and local: gardens and neighborhoods, practical skills and cash savings, relationships with people one actually sees, communities that know each other well enough to function without institutional mediation. It is what used to be called, without irony, civic life. The tendency of modern societies has been to trade resilience for efficiency in every domain — just-in-time supply chains, centralized data centers, digital payments replacing cash, search engines replacing memorized knowledge. Each trade looked rational in isolation. The aggregate effect is a civilization that is incredibly productive under normal conditions and incredibly brittle under abnormal ones. A 1920s American town could feed itself for weeks from its own hinterland and repair its own machinery. A 2026 American suburb has perhaps three days of food in local supermarkets and would be functionally helpless if the grid went down for a week. We have built a world that cannot survive its own failures. The work of building small redundancy — of knowing your neighbors, growing some of your own food, having tools you can use without the internet, preserving skills and relationships that don’t require the cloud — is not nostalgia. It is what civilizations have always been made of, and what ours is losing.
I do not know if we will come through this well. I think the odds favor muddling through rather than catastrophe, but the tails are fatter than they used to be, and the worst-case scenarios are genuinely worse than anything in modern memory because the technological and ecological stakes are higher. The specific worst case — a permanent oligarchic rule over a managed surplus population — has never actually been sustained anywhere, which is some comfort, but the transitions through it have been brutal.
What I am certain of is that fatalism is a gift to the worst possible future. Despair is what the people running the extractive arrangements want from the rest of us, because despair means we don’t organize, don’t vote, don’t talk to our neighbors, don’t raise children who know history, don’t contribute to the slow work of keeping decency alive in our corner of the world. The people who got through the twentieth century’s worst periods and left things better than they found them were not optimists. They were people who saw clearly and kept working anyway. That posture is available to anyone who wants it. It is probably what this moment calls for.
The question that defines our period is not whether humans will be needed for production. Increasingly we will not be. The question is whether we will need each other for anything else, and whether enough of us will choose to, in the absence of the old necessity that used to make the choice for us. The answer is not preordained. It is being worked out now, in every decision to build a real relationship or withdraw into a screen, every decision to invest in a place or extract from it, every decision to tell the truth about what is happening or to soften it into something more palatable. The cooperation gets chosen, or it doesn’t. If enough of us choose it, the institutions rebuild, slower and weirder than before, but real. If we don’t, the future belongs to the people who need us least.
That is the choice. It is worth naming clearly, so that those still willing to make it can find each other while there is time. Humans like you and I still decide.
Niklas Osterman