The automation paradox: who will buy when the machines take over?

In a recent address, Senator Bernie Sanders delivered a stark warning: artificial intelligence and robotics, if left unchecked, could dismantle the foundations of the modern working class. His argument extends beyond fears of technological unemployment. It raises a deeper structural question about capitalism itself: what happens to a consumer-driven economy when technology renders consumers redundant as workers?

Over the past decade, some of the world’s wealthiest figures—Elon Musk, Jeff Bezos, Larry Ellison, and Mark Zuckerberg among them—have invested hundreds of billions of dollars in AI and robotics. Their motivation is clear. Automation reduces labor costs, increases productivity, and boosts profit margins. Yet, as Sanders points out, these same efficiencies threaten to remove millions of people from the workforce. If the individuals who produce goods can no longer afford to buy them, the economic cycle that sustains growth begins to unravel.

The disappearing workforce

The current wave of automation differs from earlier industrial shifts in both scale and reach. Artificial intelligence is not limited to repetitive factory work. It now performs creative and analytical tasks once thought uniquely human: writing, coding, diagnosing illness, and even driving.

Amazon, for example, employs more than a million warehouse robots and has already laid off tens of thousands of workers since 2022. Foxconn has replaced 60,000 assembly-line employees in a single facility with machines and is moving toward fully automated production. Self-driving trucks and delivery vehicles, developed by firms such as FedEx, Walmart, and Aurora, are already operating on American highways.

A report released by the U.S. Senate Committee on Health, Education, Labor, and Pensions estimates that AI could replace nearly 100 million American jobs in the next decade, including almost half of truck drivers, two-thirds of accountants, and nearly nine in ten fast-food workers. Even if the final numbers are lower, the direction is unmistakable: automation is expanding faster than the economy’s capacity to create new forms of employment.

The implications extend beyond income loss. Work provides structure, identity, and a sense of belonging. A society that allows technology to displace millions without offering alternative forms of participation risks not only economic stagnation but social fragmentation.

The demand dilemma

The more profound issue lies in aggregate demand. Economies grow when people buy goods and services, not merely when they are produced efficiently. If automation concentrates income among a small group of capital owners while depriving the majority of wages, total consumption inevitably declines.

Initially, automation appears to enrich those who adopt it. Costs fall, profits rise, and shareholders celebrate. But over time, the same process erodes the customer base on which profitability depends. Machines may produce goods at minimal cost, but they do not purchase them. A society that replaces its consumers with robots risks self-cannibalization—an economy with immense productive capacity but insufficient demand to absorb its output.

Historically, capitalist societies have resolved this tension through redistribution. The productivity gains of the twentieth century were balanced by higher wages, reduced working hours, and the creation of welfare systems. These mechanisms kept the connection between progress and purchasing power intact. Without similar adjustments today, the next wave of automation could result not in prosperity but in chronic deflation and underemployment.

The dehumanization risk

Beyond economics lies a philosophical question. If machines perform creative, cognitive, and emotional labor better than humans, what role remains for people? Sanders and other observers warn of a potential loss of meaning. Work, for all its imperfections, provides purpose and connection. The prospect of a society in which employment becomes optional for the privileged few and unattainable for the many is both economically unsustainable and existentially corrosive.

Elon Musk’s remark that future jobs may become “a hobby” captures this tension. A world where machines perform all essential functions might promise abundance, but it could also lead to profound alienation unless wealth and purpose are broadly shared.

Rethinking the social contract

To prevent technology from eroding the social and economic fabric, Sanders argues for a reimagining of the social contract between capital and labor. He calls for a shorter workweek—specifically, thirty-two hours without loss of pay—so that workers can share in the immense productivity gains that automation delivers. Such a shift would recognize that efficiency should translate not only into higher profits but also into more time, balance, and well-being for those who generate value.

He also advocates giving workers a greater voice in corporate governance. Large companies, he argues, should allow employees to elect nearly half of their boards of directors, following the codetermination model used in Germany. If workers participate in strategic decisions, they can help shape how technology is introduced and ensure it benefits employees as well as shareholders.

Equally important is the idea of profit sharing and employee ownership. When workers have a direct stake in the enterprises they serve, the gains of automation are distributed more broadly. Rather than funneling wealth upward, this model keeps the rewards of innovation circulating within the real economy. Sanders even suggests introducing a “robot tax” on corporations that replace human labor, using the proceeds to fund retraining, universal healthcare, and income support for displaced workers. These proposals, taken together, aim to ensure that the efficiency of machines translates into social progress rather than social displacement.

The broader imperative

This challenge extends far beyond the United States. Every advanced economy faces the same crossroads. The question is whether technology will serve humanity or humanity will serve technology. Without intervention, market forces alone will direct the gains of automation toward capital owners, leaving the rest to bear the social costs.

The paradox is that even those who profit most from automation cannot escape the consequences of reduced demand. Robots may work tirelessly, but they do not consume. If income and purchasing power collapse, so too will the markets upon which corporate profits depend. Economic history shows that extreme inequality and weak consumption eventually undermine growth itself.

Addressing this imbalance requires policies that sustain the circular flow of income: shorter working hours, stronger labor protections, investment in public goods, and redistribution mechanisms that keep households solvent and engaged. These are not acts of charity but of macroeconomic necessity.

A human-centered future

Artificial intelligence and robotics could liberate humanity from drudgery and open new horizons of creativity and discovery. The danger lies not in the technology but in its ownership and purpose. If concentrated in a few hands, automation will amplify inequality and erode the social fabric. If democratized, it could usher in a new era of shared prosperity.

The true measure of progress is not how fast machines learn but how widely their benefits are distributed. Societies that manage this transition thoughtfully will emerge stronger, healthier, and more humane. Those that do not may find themselves in the paradox Sanders warns against: a world overflowing with goods that no one can afford to buy.

AI will undoubtedly transform the global economy. The question that remains is whose future it will serve—the machines that create wealth or the people for whom that wealth was meant.