AI Wealth for Workers? A Cautionary Take on Ramaswamy's Plan

Look — Vivek Ramaswamy is selling optimism: workers can build wealth in the AI era. That’s plausible. But optimism packaged as a playbook for individuals is where policy and reality split.

Start with what he gets right. AI tools really do open up new ways to monetize skills. A motivated worker can use automation to handle grunt work, move up the value chain, or spin up a niche service on the side. Treating workers as passive victims of technology is condescending and wrong; agency matters.

Here’s what nobody tells you: agency is necessary, but not sufficient, when the value chain is owned by someone else.

Ramaswamy’s column in The Wall Street Journal draws a straight line from AI tools to worker wealth. He’s right that some people will ride those waves to higher incomes. But arguing that workers broadly can replicate that ignores two structural facts: access and bargaining power. If you don’t own capital or control the platform where value is captured, you’re dependent on someone who does.

You can train a thousand people in prompt engineering; if the distribution channels and profit centers are owned by a handful of firms, most of that new value will be extracted upward. That isn’t a moral judgment — it’s how incentives and contracts work. The party that owns the rails usually owns the margins.

We’ve seen this movie before. Think about app developers in the early smartphone boom. Apple and Google did create a wave of new opportunities. A small slice of developers made real money. But because the platforms set the rules, controlled discovery, and took a cut of every transaction, the wealth concentrated at the top of the stack. Most workers never saw the kind of upside the hype promised.

Spare me the Silicon Valley–inflected bootstrapping myth: telling workers to “seize AI opportunities” assumes equal starting points. Access to the best tools, the networks to find buyers, and the capital to scale a side gig — those are unevenly distributed. Cities with dense tech ecosystems concentrate venture capital, mentorship, and high-paying clients. Rural and deindustrialized areas don’t. Telling someone in a manufacturing town to pivot to AI consulting without addressing those gaps is tone-deaf.

Where Ramaswamy’s argument does useful work is insisting that policy shouldn’t smother initiative. Workers who learn adjacent skills, automate repetitive tasks, or create niche services can capture premium pay. That’s a real channel to wealth, and there’s no need to sneer at people trying to take it.

But agency plus structural reform is what changes outcomes at scale. Tax incentives that favor capital without parallel workforce supports will skew gains to investors, not labor. You can’t tell everyone to become “AI-enabled” and then design the tax code and corporate rules so the upside accrues mainly to shareholders.

Here’s what nobody tells you: the boring design details decide whether AI feels like a ladder or a trap.

Consider two levers — portability of benefits and negotiated profit-sharing. If benefits and training credits follow the worker, not the employer, people can change jobs or start side ventures without gambling their health coverage or retirement. If companies are nudged or required to share productivity gains with employees — via profit-sharing, co-ownership models, or broader stock access — AI-driven productivity doesn’t just turn into higher executive bonuses.

Look at Costco or Publix. They’re not utopias, but both have long-standing models where employees participate more directly in the economic upside than in most of corporate America. Translate that logic to AI adoption — sharing some of the cost savings and revenue lift with the people whose jobs are being “augmented” — and you get a different trajectory for middle-skill workers than if you treat them as a line item to be optimized.

A hard truth: those kinds of structural shifts are political and operationally messy. Companies will argue that regulatory costs reduce competitiveness. And critics of profit-sharing mandates or benefits portability are right about one thing: heavy-handed rules can backfire if they ignore how markets respond. Capital can sidestep clumsy policy.

Give me a break, though, from the idea that the only choices are laissez-faire cheerleading or regulatory overreach. There’s a middle route: targeted reforms that change incentives without breaking markets. Think portable training vouchers tied to actual skill acquisition, not just seat time. Tax credits for firms that can show measurable increases in frontline pay connected to productivity gains. Rules that make it easier — not harder — for workers to form co-ops or own equity in platforms where they create value.

These aren’t fantasies; they’re design problems. Get the metrics, eligibility rules, and audit mechanisms right, and you tilt the playing field just enough that “learn AI” stops being a lottery ticket and starts being a pathway that pays off for more than a tiny elite.

Another blind spot in the column is the assumption that all industries will respond to AI the same way. Healthcare, legal services, creative work, logistics, and manufacturing have different frictions: certification requirements, regulators, capital intensity, and physical constraints. A software-based gig can scale quickly. A nurse can’t double patient load without blowing up safety. One-size-fits-all advice about “using AI to increase your influence and economic power” doesn’t survive contact with sector reality.

We should stop pretending that individual grit is a policy. Encouraging workers to learn is good; celebrating entrepreneurs is fine. But when commentary about personal hustle shows up where structural proposals should be, it turns into a cliché dressed up as guidance.

If Ramaswamy’s thesis takes off, expect a wave of AI success stories on stage — and a quieter story underneath about who wrote the rules for sharing the gains.