What becomes of humans in a world where they hold zero productive value?
Our societies invest in people because, when machines are tools rather than agents, economic activity depends on metabolic labor. Schools, culture, hospitals, universities, time off... These benefits do not stem from the benevolence of the State, but from a cold calculation: educated, healthy, and well-rested people are more productive. They create more value, lining the pockets of the powerful.
It takes 20, 25, sometimes 30 years to raise and train a human to be productive—working seven or eight hours a day, five days a week, for 35 or 45 years. It takes a fraction of a second and pennies on the dollar to spin up an AI agent that neither eats nor sleeps, and comes pre-loaded with all the knowledge and skills of the models that came before it.
The looming dystopia is easy to see. This is what Drago and Laine call "the intelligence curse." It draws on the concept of the "resource curse," which explains why countries rich in resources (oil, gas, minerals) often suffer from high poverty rates and authoritarianism (e.g., the DRC or Venezuela). Beyond buying social peace, these rentier governments have no incentive to support their populations, since State revenues depend very little on human labor. Drago and Laine note counter-examples (like Norway) that offer paths to breaking the curse. Yet, they argue the default outcome is the rise of an exclusive, powerful oligarchy of AI owners, resulting in extreme wealth concentration and equally extreme destitution for the rest. This is "techno-feudalism": a system where AI owners are the lords and the rest of us are serfs. Without a social ladder to climb, inequality becomes entrenched.
Thinking the public sector will be spared is a mistake. As noted, schools, hospitals, and universities rely on an incentive mechanism—training and caring for productive citizens—that will no longer exist. Even if we wanted to maintain a human touch in care professions (childcare, teaching, nursing), the funding simply won't be there. As Drago and Laine put it, "it's not that there won't be a demand for teachers, it's that there won't be an incentive to fund schools"5.
Nor will manual jobs be spared. In China and the US, progress in robotics has been blistering (driven by the plummeting cost of intelligence) and is set to accelerate exponentially, with AI itself contributing to robotics research.
Universal Basic Income (UBI) is often touted as the solution. Elon Musk claims there will be no more poverty, that "work will be optional," and that everyone will benefit from a "universal high income"6. When such promises come from a man who, just months ago, took a chainsaw to meager American public aid programs, we are right to be skeptical.
But the argument itself is flawed. Why would the powerful support populations that have become economically useless? Furthermore, would UBI be sent to citizens of the Philippines or Bolivia, who will face the same upheavals without owning the AI infrastructure? I highly doubt it. In my view, the more likely outcome—in the Global South, at least—is a national ban on agentic AI to protect the human economy. This would lead to a rapid regression toward subsistence economies as these nations cut themselves off from international trade.
This assumes, of course, that the powerful (the nations or companies controlling the AI) leave them alone. That is far from certain. One of the greatest checks on aggression in our societies is public outrage at the sight of coffins returning from the front. When we can send robots to war, that restraint collapses.
5 Luke Drago and Rudolf Laine, op. cit.
6 Statements made notably at the U.S.-Saudi Investment Forum, Washington, D.C., November 19, 2025.