Capitalism, Upgraded: Profits Alongside Purpose, By Design

Every October since 1982, Maria Chen has received a dividend check from the Alaska Permanent Fund — her share of the state's oil wealth. In 2024 it was $1,702. She didn't drill a well or pick an energy stock. She simply lived in Alaska while the state turned a finite resource into permanent shared wealth. Somewhere this same October, the team behind an AI company hitting $100 million in annual revenue — with fewer than 100 people — is writing no equivalent check to anyone.

Both facts are real. The question is who should benefit from that productivity gain, and by what mechanism. This isn't a critique of markets. It's a design problem about a specific, old failure mode running at a new speed.

The tension: markets work, except where they don't

Let's be fair to the conventional view, because it's largely correct. Market capitalism lifted billions from poverty and generated technologies that would have seemed miraculous a century ago. Profit is a legitimate, productive signal. Markets price transactions with an efficiency no central planner has ever matched. None of that is the diagnosis.

The diagnosis is narrower and harder to dismiss. Markets struggle with externalities — the costs others bear that never appear on anyone's balance sheet. Your smartphone's location data trains a model predicting your neighbor's movements without their consent. An automated hiring system rejects a candidate who never opted into algorithmic screening. A single AI provider becomes infrastructure for millions of businesses, and when it pivots, the dependency becomes a shared crisis. In each case, a private decision generates costs that land on people who had no voice in it.

This isn't new. In 2018, Facebook engineers warned internally that optimizing for engagement would amplify misinformation. The company shipped anyway. Revenue surged; regulatory penalties came to less than the profits they protected. AI doesn't create this misalignment between private gain and social cost. It accelerates it — and severs the connection between value creation and employment entirely. The timing makes it acute: in February 2026, Goldman Sachs chief economist Jan Hatzius noted that despite roughly $700 billion in U.S. AI infrastructure investment across 2025, AI had contributed "basically zero" to GDP growth, because most of that investment flowed to Taiwanese and Korean GDP. The dividend is real but lagged. Displacement arrives in households now; broad gains are projected for 2027 and beyond. That gap is exactly where social architecture either holds or breaks.

The reframe: architecture, not redistribution

This distinction prevents a persistent misreading, so it's worth stating precisely. Redistribution moves existing wealth from concentrated holders to dispersed recipients — and its political controversy is that it requires coercion applied after the fact. The mechanisms here do something structurally different. They design the conditions under which wealth is generated so that it disperses by default rather than concentrating.

Figure: Capitalism, upgraded — three mechanisms address control, purpose, and distribution respectively. Remove any one and the remaining two bear a load they weren't built for.

Three mechanisms, three failure modes:

• Steward-Ownership: How Founders Lock Mission Into the Architecture addresses control — who decides what the company becomes. It locks mission into governance so a firm can't be sold for extraction. Carl Zeiss has operated this way since 1889; Patagonia converted in 2022. It's a founder-level lever, encoded at founding or not at all.
• Participation Dividends: Pricing the Work the Market Can't See address purpose and identity — what people do when markets no longer need their labor. They compensate caregiving, civic work, and mutual aid that markets price at zero. A municipal-and-state lever; pilots are the right scale now.
• Public-purpose funds address distribution — who captures the equity value of productivity gains. Maria Chen's check is the working proof: every Alaskan gets an equal share, protected by a constitutional supermajority requirement against raids. A national lever, the slowest to build.

They are not interchangeable. Steward-ownership without distribution leaves mission-locked firms serving a narrow ownership base. Public-purpose funds without purpose infrastructure generate passive recipients, not contributing citizens. Participation dividends without equity stakes distribute identity without material foundation. Each solves a distinct problem; remove one and the others carry a load they weren't designed for.

How it works: the business case, not just the ethics

The strongest reason for business leaders to take this seriously isn't moral. It's that excessive concentration creates business risk, not only ethical risk.

Companies that shed workers during the displacement-to-dividend lag reduce the consumer base sustaining their own revenue — a mechanism that accelerates rather than corrects. Larry Fink put the historical pattern bluntly in his March 2026 BlackRock letter: "Since 1989, a dollar in the U.S. stock market has grown more than 15 times the value of a dollar tied to median wages. Now AI threatens to repeat that pattern at an even larger scale." Jamie Dimon was blunter still at Davos in January 2026: "You can't lay off 2 million truckers tomorrow." Luxury markets survived the 2008 recession; consumer brands serving the middle didn't recover for eight years.

And the firms that built stewardship into their structure tend to outcompete their extractive equivalents over long horizons — not because their leaders were more virtuous, but because the architecture made short-term extraction for long-term cost legally difficult. Patagonia's structure allowed 15-year R&D bets in recycled materials; its Worn Wear resale program now runs customer lifetime value at roughly 3× the industry standard. Novo Nordisk's foundation governance protected GLP-1 research for 18 years before Ozempic — shareholders demanding quarterly returns would have killed it at year seven. The apparent trade-off between profit and stewardship is real on a quarterly horizon. On a 15-year horizon it largely dissolves. The conflict isn't profitability versus purpose. It's which period you're optimizing for — and in frontier markets where compounding long bets determine winners, that choice is the strategy.

What to do

1. Know which lever is yours. A founder has a steward-ownership decision this year. A mayor has a participation-dividend decision. A legislator has a public-purpose-fund decision. They're not the same decision, and conflating them produces paralysis.
2. Act inside the window. All three mechanisms work best before concentration hardens. Steward-ownership at founding is far more durable than retrofitting it onto a company already structured around extraction. Funds that acquire stakes at $10 billion valuations capture more than those buying in at $1 trillion.
3. Treat the multilateral problem honestly. Tax arbitrage and the geopolitical race are real constraints — a single jurisdiction adopting stewardship while competitors offer havens chooses disadvantage. The answer isn't to abandon the architecture but to build it the way OECD Pillar Two was built: simultaneously, across jurisdictions. The mechanisms are necessary but not sufficient on their own.

The principle underneath all three is that architecture precedes virtue. These mechanisms aren't appeals to the generosity of founders. They're structural designs that make extraction difficult and stewardship self-reinforcing — regardless of who holds the reins in 2040. Governance structures aren't constraints on prosperity. They're conditions for it. That's what "capitalism, upgraded" means: not profits versus purpose, but profits alongside purpose, encoded into the structure of the enterprise itself.