Ramaswamy's AI Wealth Pitch Overlooks Shared Ownership and Safeguards

Vivek Ramaswamy says workers can build wealth in the AI era in a Wall Street Journal opinion piece. Here’s the thing: that optimism is partly right — AI will create value — but it underestimates how value is captured.

Ramaswamy’s basic story is familiar: new tech, new opportunities, new ladders to climb. Fine. I’ll be honest — I want that story to hold. But if you’ve spent any time reading Silicon Valley term sheets instead of launch blogs, you learn a less romantic plotline. Software firms accumulate intellectual property, platforms extract rents, and capital owners take the upside. Workers usually get what’s left after the stock grants, licensing deals and platform fees are spoken for.

Look at the rise of cloud and app ecosystems. The tools absolutely empowered developers and small teams. They also funneled enormous, reliable returns to the platforms that controlled billing, distribution and data. That isn’t villainy; it’s incentives plus contracts. And those same mechanics will govern who wins in AI.

That’s why the column’s use of “workers” as a single, cohesive class feels off-key. An engineer with equity at a successful AI startup might build generational wealth. A delivery driver replaced by autonomous routing software won’t. A barista who spends her shift training a customer-service chatbot doesn’t get a residual every time that model answers a query. Unless she owns a slice of the model — or the company voluntarily cuts her in — her contribution dissolves into corporate margin.

Ramaswamy gestures toward “opportunity” without staying long enough on the boring stuff: who holds equity, who sets prices, who owns the models and datasets, and who drafts the employment contracts. Those levers determine where AI’s surplus sticks.

This isn’t abstract. If AI just lowers the marginal cost of producing digital services, the surplus tends to flow to whoever owns the choke points: the foundation models, the data pipelines, the app stores, the interfaces glued to millions of users. That’s how certain tech hubs end up with outsized stock-option wealth while other regions get gig work and volatility.

William Gibson sketched a version of this decades ago: a dazzling matrix of information, with power quietly saturating a handful of corporate intelligences while hackers and hustlers scrape along the edges. The magic trick in that world wasn’t the tech; it was who owned the black boxes.

Now, to Ramaswamy’s credit, he doesn’t just say “adapt” and walk away. He leans hard on skill-building. That matters. Skills are portable currency; they fuel job hops, career pivots, and occasionally, those “I quit to build a startup” LinkedIn posts.

But training alone doesn’t grant ownership.

Employers already find ways to capture the upside of investing in worker skills. They use credentials as gatekeeping tools. They deploy restrictive IP clauses. They monitor productivity with software that turns every keystroke into a performance metric. When firms control the training budget, they usually control the bargaining environment around that training.

Policy choices decide whether public or employer-backed training funds buy people a genuine stake in AI’s gains or simply make them more competitive in a labor market where they still don’t own anything.

So what would make his thesis feel less like a pep talk and more like a plan?

Start with ownership mechanisms that meet workers where they actually are. Employee stock-ownership plans and broad-based equity aren’t just for early-stage engineers; they can be structured for support staff, operations teams and, yes, frontline workers. Profit-sharing schemes can tie payouts not just to quarterly earnings but to measured productivity gains from AI deployments, so the people who absorb the disruption see more than a “thanks for your flexibility” email.

Then address portability. If a warehouse worker spends three years being trained to use AI-powered logistics systems, those credentials shouldn’t turn to dust the moment she changes employers. Think of training as an asset that belongs to the worker first, with systems that let her carry those verified skills — and sometimes even a slice of upside from the tools she helped tune — to the next job.

Skeptics will counter that markets already handle this via entrepreneurship. New tools will spawn new small businesses, creators, cottage industries. And that’s partly true. We’ll see solo operators using AI to punch way above their weight, automating back-office work and competing with much larger firms.

Sure, but market-led “democratization” of ownership tends to be slow and patchy. For every breakout small business that uses AI to thrive, there are dozens of users whose data and creativity feed the models for free while platforms collect the compounding returns. Network effects and platform dominance are powerful centripetal forces; without explicit guardrails, they pull value back to the center.

History here is less about doom than pattern recognition. When railroads standardized gauges and schedules, they created huge productivity gains — and also enormous fortunes for those who owned the tracks and rights-of-way. Securities laws and corporate forms came later, partly as a response, to widen who could invest and how. Regulation didn’t kill the railroads; it set the terms for who could ride and who could own stock in the ride.

The same logic can apply to AI. Disclosure rules around model ownership won’t stop innovation, but they’d clarify which entities actually control the engines of value. Tax incentives that reward companies for broad-based equity or profit-sharing can tilt boardroom decisions without micromanaging products. Portability requirements for training credits can turn “upskilling” from a corporate PR line into a tangible, bankable asset for workers.

Ramaswamy is asking the right headline question: how do workers build wealth in the AI era? The answer hinges less on inspiration and more on how we wire ownership, contracts and bargaining power into the stack.

If those details don’t change, AI will feel less like a fresh start and more like a faster sequel to the same old wealth script Ramaswamy claims he wants to rewrite.