Why AI Won't Owe Workers Wealth—They Must Build It

Ramaswamy says workers can build wealth in the AI era — and he's right to force the conversation about ownership. Here's the thing: the promise of ownership is emotionally and politically powerful, but rhetoric doesn’t pay rent. The Wall Street Journal column taps into a real anxiety about who benefits from AI, then races past the hard part: how regular workers actually gain equity in the systems that may displace their jobs.

Let’s start with what the article gets right: it plants a flag on democratic ownership. Workers should capture some of the upside of automation instead of just eating the disruption. That’s a welcome contrast to the usual “learn to code and pray” advice. It echoes old-school arguments about broad-based capitalism and employee capitalism that have floated around policy circles for years.

But “workers should own more” is not a strategy. It’s a vibe.

Who organizes that ownership? Who sets the terms? Who makes sure workers don’t get wiped out when a company restructures, pivots, or quietly buries the AI project they were “equity partners” in? Those are design questions, not talking points, and they matter more than optimistic slogans about shared prosperity.

Look, companies from Microsoft to Amazon are extracting enormous value from AI systems built on data, talent, and infrastructure that most people will never directly touch. Why shouldn’t workers get a slice? They should. But getting that slice requires institutions that translate aspiration into actual cap tables, contracts, and protections.

This is where geography and structure crash the party. AI’s gains are clustering around a handful of platforms and cities — the usual suspects on the coasts and a few research-heavy metros. Talent and capital don’t just float gently across the country; they agglomerate. So most workers — the service staff in smaller cities, the line workers in less glamorous manufacturing towns — are structurally distant from the companies and funds minting AI wealth.

Hand someone an equity grant in a hot startup and, without liquidity, financial education, and guardrails against volatility, you’ve mostly handed them a lottery ticket. Or, worse, a lottery ticket they’re pressured to hold just as the market rolls over. That’s not broad-based capitalism; that’s exposure without power.

Sector matters, too. Some workers will naturally become owner-operators in this wave: software developers building AI tools, data scientists tuning models, small entrepreneurs layering products on top of existing APIs. They’re close to the technology and the value chain.

Others — health aides, truck drivers, retail clerks — will see their tasks reconfigured by AI scheduling, routing, and monitoring systems without any corresponding path to equity. You don’t magically get a stake in an AI-enabled logistics platform just because your route got optimized.

That asymmetry is a major blind spot in the ownership narrative. The article gestures toward “workers” as a single category, but AI doesn’t hit a radiologist, a warehouse picker, and a school aide in the same way. Policy prescriptions shouldn’t treat them like a monolith.

I’ll be honest: the magic-bullet idea that everyone can simply pivot into high-value AI work is fantasy. This isn’t some Charles Stross Accelerando scenario where a quick software patch upgrades you from cashier to machine-learning engineer between breakfast and lunch. Real life has prerequisites: education, childcare, transportation, housing, local institutions. Those aren’t footnotes; they’re the main show.

If you take the ownership idea seriously, you end up back in the world of unsexy but powerful mechanisms: Employee Stock Ownership Plans (ESOPs), genuine profit-sharing, worker cooperatives tied to real revenue, not branding exercises. These tools already exist. They just haven’t been updated for a world where AI value often comes from platforms, data networks, and software layers several steps removed from the frontline worker.

And then there’s the US retirement system as a quiet counterexample. For all its flaws, broad-based retirement accounts have turned millions of workers into indirect owners of the corporate economy. Not founders. Not VCs. But owners, via mutual funds and index funds. You want a template for “democratic” AI ownership? Start with how pension funds and retirement plans allocate capital, and ask why that capital is rarely steered toward worker-inclusive AI ventures.

Policy can push this further. You could imagine incentives for companies that tie a meaningful slice of AI-driven productivity gains to employee profit-sharing pools. Or regional investment vehicles where public funds, unions, and local institutions back AI-related businesses while locking in an ownership floor for employees. That’s messier than a catchy op-ed, but that’s where the real power sits.

The likely counter-argument is predictable: free markets will create new ownership opportunities on their own. New firms, new platforms, new gig-style arrangements where workers can “participate” in AI ecosystems. Yeah, no — markets do create wealth, but they also have a well-documented habit of concentrating it without guardrails. Tech history looks like a sine wave of concentration and backlash: railroads, telecom, early internet, cloud platforms. Each time, we get a flurry of innovation, then consolidation, then a scramble to retrofit rules.

So sure, let markets innovate. But if the goal is workers actually building wealth, not just surviving the transition, you need rules and norms that push ownership outward instead of letting it condense at the top of the stack. That’s less about heroic founders gifting shares to the people and more about tax codes, securities rules that make small-stake ownership safer and more liquid, and corporate governance that treats workers as residual claimants, not just a cost line.

I keep thinking about a robotics demo I saw where a humanoid fetched coffee while executives on stage talked about “freeing humans for higher-value work.” Nobody on that stage was the barista or the warehouse worker whose job might quietly shrink five years later. That’s the core tension: AI is built and showcased by the people who already own plenty, and discussed as an opportunity for the people who don’t.

If Ramaswamy’s column pushes that crowd to start talking about actual ownership structures instead of hand-waving about “worker wealth,” it will have done more than its word count suggests. My bet: the loudest AI winners will adopt some form of profit-sharing and equity for workers — not out of pure altruism, but because the politics of the AI era will make “ownerless” disruption a harder and harder sell.