The Economy Runs on Paychecks. AI Is About to Eliminate Millions of Them.

Both the tech right and the progressive left are failing to prepare. But there’s a model that already works.

Why do companies pay anybody to work there?

A company would love to keep every dollar of revenue. The only reason it hands some of those dollars to you is that it can’t get what it needs any other way. Your time is finite. Your skills took years to build. There aren’t enough people who can do what you do. You have leverage, and leverage gets paid.

We never think about this. It’s like gravity.

AI is about to turn off the gravity. Not because it’s smarter than you, but because it’s cheaper, faster, and never needs to sleep. I studied computer science in college and have spent the last year doing a deep dive into AI, so I don’t say this as someone who’s afraid of the technology. I say it as someone who uses it every day and can see what’s coming. For white-collar work across accounting, legal research, marketing, software development, and project management, “good enough and endlessly available” is already here.

Since 2022, tech companies worldwide have laid off more than 700,000 workers. In just the last week, Jack Dorsey cut nearly half of Block’s workforce and said explicitly that AI made them unnecessary. His stock surged 24%. Then he predicted that most companies would follow within a year. Microsoft’s AI chief, Mustafa Suleyman, told the Financial Times that most white-collar tasks will be “fully automated by an AI within the next 12 to 18 months.”

Every one of those lost paychecks ripples outward: payroll taxes that fund Social Security, income taxes that fund schools, mortgages that prop up banks, consumer spending that keeps Main Street alive. Multiply that by millions and the entire circulatory system of the American economy starts to shut down. Blue-collar workers lose their customers. Capital loses its returns. Social Security, already heading toward insolvency around 2033, accelerates toward the cliff. There is no version of this where one part of the economy collapses and the rest keeps humming.

But there is a way through it. Before we get there, though, we need to understand why the two groups with the most power to act are both making it worse.

---

Two Failures

The tech oligarchs see the problem coming. Sam Altman talks about universal basic income. Elon Musk muses about a world where no one needs to work. Marc Andreessen writes manifestos about abundance for all. Now watch what they do: they’re lobbying to gut Social Security and Medicare, buying up AI infrastructure so a handful of companies control it, and arguing that government oversight would hold humanity back. They want to tear down the current system beforethey’ve built the replacement.

The logic is straightforward: every dollar the government collects in taxes is a dollar that can’t be invested in the AI buildout. When they argue for cutting public programs, they’re clearing the field for their own capital needs. But “abundance is coming” is not going to cover anyone’s mortgage.

That kind of suffering is easy to overlook from a private jet.

Meanwhile, the left has so far failed to articulate a compelling alternate vision. A significant chunk of the left wants to block AI entirely, which won’t stop it from being built — it’ll just be built by people with less interest in fair outcomes. The right has a coherent economic story: free markets, low taxes, growth that lifts all boats. It’s wrong in important ways, but it’s a story. The left’s counter-story is too often just a catalog of things we oppose. A list of problems is not a vision. And you can’t win the most important economic fight of the century without one.

So let me offer one.

---

The Alaska Answer

In 1976, Alaska faced a version of the problem we’re about to face. The state was sitting on enormous natural resource wealth, and the question was straightforward: who benefits? The oil companies that extract it? The government that taxes it? Or the people who live there?

Alaska chose everyone. They amended their constitution to create a permanent fund, invested the oil revenue, and starting in 1982 began sending every resident an equal annual dividend. The program isn’t perfect — the dividend amount has become a perennial political fight, and since 2018 the fund has also been tapped to cover more than half of state government operations. The 2025 dividend was $1,000, the lowest in inflation-adjusted history. But the core principle has survived over four decades of political pressure: when the land produces wealth, the people who live on it get a share, not as charity, but as a right built into the structure. The fund now stands at more than $83 billion.

AI is about to become the most productive resource on earth. So who gets the share?

Right now the answer is a very small number of companies and their investors. But AI didn’t emerge from nothing. It was built on publicly funded research, trained on data generated by billions of people, developed by workers educated in public universities, and deployed over an internet that began as a government project. The public has a legitimate claim to a share of the wealth this technology produces. A national fund built on equity stakes in AI-driven companies would formalize that claim. The machines generate the wealth. The fund captures a share. Every American gets a dividend.

That’s the foundation. But a dividend alone won’t be enough if the rest of the economy is collapsing around it. The tax base has to follow the value: right now we fund nearly everything by taxing paychecks, and if AI eliminates the paychecks, the funding vanishes. Tax the company that replaces 500 workers and triples its output, because that value is real and the current code barely touches it. Build a real bridge for the people caught in the transition, because the current unemployment system is useless for people whose entire profession just evaporated. Put people to work on things that actually matter, because we have a massive aging population without enough caregivers, infrastructure that’s falling apart, and schools that are desperate for teachers — the work is there, the market just won’t fund it, and a WPA-scale public employment program would absorb displaced workers while getting critical things done.

These are the mechanics. Every one has precedent. None is utopian. And yes, they require a government that works better than the one we have now. Building that capacity is itself part of the project, and anyone who says we can’t is really saying we shouldn’t try. But these are only the machinery of a larger idea, and the larger idea is what actually matters.

---

What We’re Building Toward

For most of human history, economies have been organized around scarcity. There isn’t enough, so we compete. Your worth is measured by your productivity, your salary, your ability to out-hustle the next person for a finite pool of resources. Every institution we’ve built reflects this: employment as the gateway to healthcare, retirement tied to decades of labor, social standing measured in dollars.

If AI delivers on even a fraction of what its builders promise, that scarcity starts to loosen. And the question we need to be asking is: what do we build in its place?

A purpose-driven economy would start from a different premise. Instead of asking “how do we allocate scarce resources?” it would ask “how do we help people live meaningful lives?”

It means a 55-year-old teacher who’s been scraping by on an inadequate salary can actually do the job our society claims to value. It means the woman who’s been caring for her aging mother full-time, for free, because nobody else will, gets recognized and compensated for doing some of the most essential work in the economy. It means a kid from a working-class family can pursue architecture or nursing or music without taking on a decade of debt, because education is treated as an investment in people rather than a consumer product. It means someone who wants to rebuild wetlands or restore affordable housing or mentor teenagers in underserved neighborhoods can do that work and still feed their family, because we’ve decided that what matters to a community is worth funding even when the profit margins are thin.

This is achievable. It requires the same thing Alaska required: a decision to share the wealth that a resource generates, and institutions strong enough to make that decision stick. The resource has changed — it’s AI instead of oil — but the principle is identical.

The transition won’t be painless. But consider the alternative: an economy where a shrinking number of people own the machines that do the work, and everyone else depends on their goodwill for survival. We spent centuries climbing out of the last version of that arrangement. We called it feudalism.

Progressives need to tell this story. Neither a list of grievances nor a set of policies, but rather a story about where we’re going and why it’s worth getting there. People don’t organize around complaints for long; they organize around a destination. The destination is an economy built around human purpose and dignity, where the question isn’t “how do I survive?” but “what do I want to contribute?” The path runs through shared ownership, public investment in work that matters, and a tax system that follows value wherever it’s generated.

None of that is utopian. None of it requires technology we don’t have. What it requires is political will, built before a handful of people quietly lock up the most powerful economic engine ever built.