Why Ramaswamy's AI wealth plan won't fix wages

Vivek Ramaswamy argues, in a Wall Street Journal opinion column, that workers can build wealth in the AI era. Fine. Here's the thing: the headline makes a promise — a how-to — and the piece, at least in the excerpt I saw, offers no hard data and no real mechanics. I'll be honest: optimism without mechanisms is like bringing a laser pointer to a robot uprising — entertaining, but not especially useful. (Also, I saw a smart toaster at CES last year that insisted it was a lifestyle brand; so maybe I'm jaded.)

But let’s be fair to the headline for a second. The intuition that AI doesn’t have to be purely extractive is right. We’ve had moments where technology did broaden wealth: think of employees at companies like Microsoft or Google whose stock grants turned middle-class salaries into nest eggs. That’s the dream Ramaswamy is gesturing at — harness the big wave, don’t get crushed by it. If you squint, you can almost see a blueprint for a worker-centered AI boom hiding in there. Almost.

Then you look for the actual blueprint. And it’s just…vibes.

Ownership, not just optimism
Ramaswamy’s title taps a familiar strain of American optimism — work hard, adapt, prosper — which is appealing. But optimism alone doesn't create claimable assets. Wealth is concentrated where ownership lives: equity, stakes in platforms, intellectual property. If the claim is that workers will build wealth by reskilling into higher-paying AI roles, that’s a labor-market story; if the claim is that AI will create new entrepreneurial opportunities, that’s a capital-access story. You can’t have both answers without explaining how ordinary workers move from wage earners to asset holders when the engines of value are increasingly algorithms and centralized datasets controlled by a few big firms. The headline promises a bridge; the argument, as far as we can see it, just points across the river and says “swim harder.”

Look, startups hand out options. Big tech hands out RSUs. Neither franchise scales easily to a barista in Cleveland or a security guard in Phoenix. The omission matters because it shifts the burden from structural change to individual hustle — vintage Silicon Valley logic that sounds good in a pitch deck and less good when payrolls are sticky and housing costs are not. I like his can-do tone; I just want the road map to be visible, not implied. Without that, “workers can build wealth” becomes a story about exceptional outliers, not a pattern most people can realistically follow.

Who wins when code replaces craft?
AI is an industrial force — think factories of the mid-19th century, but with fewer chimneys and more server racks. History doesn't give us a clean moral lesson; it gives us messy redistribution. Sure, new tasks emerge. But returns often accrue to capital owners first. If Ramaswamy’s argument sidelines that reality, then it’s a partial diagnosis. Industrial transitions tend to look generous from the balcony and brutal from the factory floor.

I’ll give him credit: the idea that workers could share in AI’s upside isn’t crazy. Profit-sharing, employee ownership models, and cooperative structures have precedents. Germany’s codetermination model, employee stock plans at big tech firms, even smaller experiments in platform co-ops — none of this is science fiction. But the column excerpt I read doesn’t wrestle with the political economy: how do you convince large platform companies, whose business models rest on data monopolies, to open ownership to vast, unorganized workforces? Or how do you redesign incentives when automation reduces the marginal productivity of many wage roles? These are structural questions; slogans won’t do. You need governance, bargaining power, and rules — not just plucky individuals waving Coursera certificates.

Skills are necessary but not sufficient
A common refrain — and likely part of Ramaswamy's case, given the headline — is that reskilling will let workers climb into higher-value AI work. Skills matter. No argument there. But skills without portable access to capital, time, and safety nets are often just better ways to be exploited. Training programs that assume people can drop their lives, bail on caregiving, and spend a year in a boot camp are written by people who haven't balanced a mortgage, two kids, and a day job while networking on LinkedIn.

I'll be honest: I want to believe in human adaptability. Science fiction has always flirted with AI uplifting ordinary folks — Philip K. Dick did it as irony, Cordwainer Smith did it as myth. But real-world transitions need institutions that cushion the fall and share gains; otherwise, “skills” becomes a euphemism for “make yourself employable on someone else’s terms.” We’ve already seen hints in today’s AI boom: the prompt engineer with a fancy title and a contractor behind them doing data labeling for low pay. Same system, shinier job descriptions.

The missing middle: mechanisms and power
Here’s where the argument really wobbles. If workers are going to build wealth, you need a theory of power, not just a theory of learning. Who sets the terms for data ownership? Who decides whether AI productivity gains flow into wages, profits, or share buybacks? Who gets to build on top of these models without paying gatekeeper rents? This is the unglamorous plumbing — contracts, unions, regulation, corporate boards — that turns “you can build wealth” from a motivational poster into something like a plan.

There are plausible levers: broad-based equity plans tied to AI-driven productivity, data-dividend style arrangements where communities that generate training data share in upside, tax incentives for companies that expand ownership to non-traditional workers. None of that requires a techno-utopian leap of faith. It does require acknowledging that markets don’t magically deliver inclusive ownership without a nudge, or several.

Counter-argument — and why it doesn't land
Okay, counterpoint: markets will adjust; firms will create value and distribute some of it through higher wages and new firm-creation opportunities. Sure, but history shows market adjustments are uneven and slow; they don’t automatically produce broad-based wealth. You can point to pockets of dynamism and say, “see?” — and you'd be right about pockets. But if your promise is broad-based worker wealth, pockets are not the same thing. Yeah, no — cheerleading for market magic without institutional or policy levers leaves most people out.

A practical ask: if Ramaswamy wants people to believe workers can build wealth, tell us how ownership expands — whether through policy nudges, platform incentives, or new corporate norms. Don’t just sell hope. The headline promises a map; what workers are getting so far feels closer to a motivational poster taped to the factory door of the AI era.