Worker Ownership Is the Key to AI-Era Wealth

Vivek Ramaswamy argues in The Wall Street Journal that workers can build wealth in the AI era; here's the thing — his playbook reads like a 21st-century self-help manual sold at a venture conference. He praises individual hustle, entrepreneurial pivoting and skill stacking as the route to prosperity. Fine. But that’s only half the map; the other half — the terrain people actually live in — is full of institutions, power imbalances and legacy contracts that don’t just bend to willpower.

Let’s give him his due first. Ramaswamy is right about one enduring truth: technology opens doors that weren’t there before. Whole careers exist today that would have sounded like cyberpunk fan fiction when the commercial web was young. Telling workers to experiment with AI tools, think like entrepreneurs and avoid sleepwalking through disruption beats the usual “learn to code” platitudes.

But once you zoom out from the individual, the story changes.

Who gets paid for automation?

Ramaswamy’s column mostly treats AI as a tool that ambitious workers can pick up and run with. That’s partially true. It’s also true that companies like Google, Microsoft and Amazon capture enormous value by owning platforms, data and distribution — and by deciding who accesses expensive compute and on what terms. The history lesson here is ugly but useful: earlier waves of mechanization boosted productivity while concentrating gains among those who owned the machines, not those who fed them.

Charles Stross’s Accelerando pushes that logic to an extreme, imagining capital compounding so fast that humans are basically spectators. Ramaswamy isn’t arguing for that dystopia, but the pattern he underplays — technology magnifying existing power — is not science fiction.

His prescriptions — reskill, specialize, use AI to augment your role — are sensible for a subset of workers. They assume time, savings and a safety net to experiment. Many workers don’t have any of those. Winner-take-most dynamics mean the first movers and platform owners can soak up outsized returns while second-tier specialists scramble in their wake.

Yeah, no, this isn’t just about grit and “mindset.”

The column gestures at mobility and initiative, but it largely ignores market structures that make mobility costly: noncompete clauses that shadow you to your next job, data monopolies that lock value inside a handful of walled gardens, and gatekeeper APIs that charge for access to the very tools workers are told to master. Those institutional facts shape who can actually “build wealth” and who ends up renting their own productivity back from platforms.

Not enough talk about shared infrastructure

There’s a big blind spot in leaving wealth creation to individual agency: infrastructure matters. Access to reliable training, affordable compute and distribution channels isn’t evenly spread across cities, industries or demographics. Saying “double down on your unique skills” without asking who gets access to the amplifiers of uniqueness is like telling sailors to row harder while their competitors own steamships.

We’ve seen a softer version of this movie in the app economy. Apple’s App Store created millionaires — and also a long tail of developers who discovered that 30% fees, algorithmic promotion and platform risk made “just build an app” more slogan than strategy. The opportunity was real; so were the gatekeepers.

Policy choices sit quietly underneath all of this. They help determine whether AI augments a broad middle class or mainly fattens shareholder returns. This isn’t a plea for a regulatory chokehold; it’s a practical observation. Public investments in retraining, portable benefits and even smaller steps like ensuring workers can port their data and use it to demonstrate their productivity are the scaffolding that makes Ramaswamy’s advice realistic for more than the already-secure.

I’ll be honest — tech evangelists love the inspirational anecdote because it’s tidy. The messy work of building training pipelines, funding transition support and untangling labor law is less sexy, but that’s where scale actually happens.

Who shoulders the transition costs?

Ramaswamy emphasizes individual entrepreneurship, but he skims past who bears the downside risk when automation displaces jobs. That omission isn’t academic; it’s the core distributional question. If firms automate and keep the upside while workers absorb retraining costs, income volatility and career detours, you don’t get a society of scrappy owner-operators — you get stronger firms and a more fragile labor market.

Pull this thread and more concrete levers appear: antitrust enforcement to curb platform gatekeeping; incentive structures that encourage employee ownership models; subsidies or shared infrastructure for compute so AI experimentation isn’t restricted to corporations and elite universities. Think of it less like handing out lottery tickets and more like building public transit — everyone’s individual choices suddenly go farther.

There’s also a cultural piece Ramaswamy glides past. When every worker is told to act like a founder, the quiet, essential jobs — nursing, teaching, public service — implicitly get framed as failure to “play the AI game” properly. That’s not just unfair; it’s strategically dumb. An AI transition that treats half the workforce as collateral damage and the other half as startup CEOs is going to hit political resistance long before it hits economic maturity.

Counter-argument and a measured pushback

A reasonable rejoinder is: market incentives and opportunity have always produced winners without heavy-handed policy; entrepreneurs will find pathways, and we shouldn’t chain innovation to regulatory fears. Sure, but markets don’t fix distributional problems quickly; they tend to crystallize them. If you start the race on uneven ground, “let the best runner win” just blesses the starting positions.

The trick isn’t to stop entrepreneurs; it’s to change the rules so entrepreneurship doesn’t require preexisting wealth, privileged geography or a lucky network. That means lowering the cost of experimentation and the punishment for failure.

A concrete lever is public-private partnerships that underwrite compute credits and standardized credentialing, making it feasible for more people to tinker with AI without betting their rent. Another is clearing away frictions that lock workers into precarious arrangements — noncompetes, opaque gig ratings, data lock-in that prevents people from carrying their work history and performance signals to the next platform.

These are boring fights in courtrooms and committee hearings, not triumphant TED talks. They also define whether Ramaswamy’s vision scales beyond a fortunate minority who already sit close to capital and code.

I’ve been covering Silicon Valley long enough to watch each new wave of tech arrive draped in promises of shared prosperity, only to discover that the default setting routes most gains uphill. Ramaswamy is right that workers can build wealth in the AI era — but unless we tune the system around them, far more of that wealth will show up on corporate balance sheets than in household bank accounts.