AI Wealth for Workers Needs Real Protections

Vivek Ramaswamy says workers can build wealth in the AI era. As a counter to “robots will take all our jobs,” that’s useful. But once you get past the headline optimism, his argument treats capital, corporate power and policy as scenery instead of the main plot. That’s the miss.

There’s a version of his thesis that’s right: workers absolutely can benefit from AI. We’ve seen technology waves lift some boats before. Microsoft’s rise minted plenty of millionaire employees. So did Google’s. Early staff with real equity exposure turned productivity gains into personal balance sheets.

But those examples share one feature Ramaswamy glides past: meaningful ownership.

Free-market pep talk vs. who actually owns the rails

His script is familiar: smarter tools plus hustle equals upside for workers willing to adapt. The column talks about initiative, mindset and reskilling.

What it barely touches is the layer that actually converts AI output into money: control of models, data, distribution and customer relationships. Those aren’t neutral “inputs.” They’re the tollbooths.

Workers can absolutely learn to prompt, code, consult or spin up micro-businesses. That’s positive. Let’s be real, though: if they do it on top of platforms they don’t own, with capital they don’t control, most of the upside flows back to the platforms. You keep the task risk and maybe some fees; they keep the equity.

The math doesn't lie: who owns the appreciating asset captures the compounding. If AI-rich firms hold the equity while everyone else rents the tools, you don’t get broad-based wealth. You get a more “efficient” version of the same hierarchy.

I watched this up close on a trading floor. When new software shaved costs or expanded trading capacity, the story was always framed as “good for everyone.” But the durable gains showed up in the stock price, not in the base pay of the people actually using the tools. Incentive plans were engineered to look participatory while preserving tight control at the top.

Ramaswamy nods to ownership. He doesn’t grapple with how often it’s purely symbolic.

Training is not ownership

He leans hard on reskilling. That’s fine as far as it goes: workers do need AI fluency.

But training is a human-capital story. Wealth is a financial-capital story.

You can have an economy full of highly trained prompt engineers and AI-assisted marketers and still concentrate wealth, if those workers don’t hold meaningful claims on the businesses whose margins they expand. Upskilling raises your wage ceiling a bit; equity tied to real cash flows is what compounds over time.

There’s a big policy and governance gap here. Employee stock plans exist, but design decides whether they’re wealth-building or theater. If options are tiny, back-loaded, or constantly diluted by new rounds and executive grants, you don’t get life-changing stakes. You get a talking point in a company town hall.

Tax treatment matters too. Rules that favor certain types of investors or equity holders effectively decide who can participate in upside and who just pays ordinary income tax on their labor. If those rules stay skewed, AI doesn’t fix it. It accelerates it.

Geography, friction and the “anyone can do this” myth

Ramaswamy writes as if opportunity moves at the speed of software. It doesn’t.

AI’s benefits will cluster where capital, talent, and regulatory tolerance already sit. Think urban tech corridors, not rural service towns. A restaurant manager in a small city can use AI tools to market better or manage inventory, but that’s incremental efficiency, not a straight line to asset ownership or venture backing.

The column’s “workers can build wealth” frame assumes they can just plug into high-growth ecosystems if they try hard enough. That glosses over housing costs, family obligations, immigration barriers and plain old geography. Aspirational, yes. Frictionless, no.

Tools help. They don’t rewrite power.

There is a fair counterpoint: AI tools lower startup costs. You can build a basic product, run marketing experiments, do bookkeeping and outreach with software that used to require paid specialists. That does open doors.

But tools democratize attempts, not outcomes.

To turn a clever AI-enabled project into actual wealth, you still need paying customers, durable distribution and, often, capital for scale. That’s where incumbents—from cloud providers to marketplaces to payment rails—collect tolls and data.

History here is instructive. The internet “democratized publishing.” Yes, blogs and independent media exploded. Then a few platforms consolidated distribution and ad economics, and suddenly most writers were chasing algorithm changes while a handful of firms banked the serious returns. AI is on track to rhyme with that story unless ownership structures diverge sharply.

Capital structures are the real story

If you take Ramaswamy’s optimism seriously, you have to look past individual hustle and straight into balance sheets and cap tables.

Do workers get meaningful equity with sensible vesting? Do small businesses get financing that doesn’t saddle them with lopsided risk? Do ordinary savers get channels to own pieces of AI-driven private companies, instead of watching gains stay bottled up until late-stage listings?

Those are design questions, not attitude questions.

Ramaswamy’s column is useful as a reminder that AI doesn’t automatically doom workers. But unless the institutions that define who owns what are part of the plan, his argument reads less like a roadmap and more like a motivational poster taped to the server rack.