Imagine waking up and MANGO is gone. Not a corporate bankruptcy, not a regulator’s victory lap: Microsoft, Anthropic, Nvidia, Google, and OpenAI simply no longer organize the default path through computing. The servers still work. The researchers still exist. The knowledge has not vanished. What disappears is the assumption that a handful of firms must own the interfaces, capital, chips, models, distribution, and future.
This is a thought experiment, not a prediction. Our MANGO game caricatures the extractive loop: acquire data, acquire compute, acquire rivals, grow the moat. This essay asks what to build instead—not a smaller copy of the same empire, but companies that grow exponentially in capability while growing organically in ownership, governance, and local relationships.
*The cooperative employment estimate is from CICOPA’s 2017 Cooperatives and Employment: A Global Report, using partial data; it is a floor, not a census. The OECD figures use different definitions and should not be read as a single causal measurement.
The gap: exponential capability, monolithic ownership
There is nothing inherently wrong with scale. A large firm can fund expensive research, maintain reliable infrastructure, and ship a product globally. The problem begins when scale in one dimension becomes ownership of every dimension. A company that controls the model, the cloud, the chip supply, the app store, the identity layer, and the financing channel can turn technical advantage into a self-reinforcing political and economic position.
The evidence is more nuanced than “big is bad.” Autor, Dorn, Katz, Patterson, and Van Reenen’s superstar firms paper finds that concentration can accompany higher productivity, while also documenting a shift of sales toward firms with higher markups and lower labor shares. Bessen finds a strong relationship between proprietary IT investment and rising concentration. The defensible conclusion is not that every large company is abusive; it is that private infrastructure and winner-take-most dynamics can widen the gap between frontier firms and everyone else.
What “grow sideways” means
Exponential growth usually describes revenue, users, or valuation. We need a second curve: the number of people who can meaningfully own, govern, adapt, and benefit from the system. A next-generation company grows upward in capability and outward in agency.
Automate the repeatable work and invest in tools that make a small team unusually effective.
Publish interfaces, portability guarantees, and replaceable components so customers are not trapped.
Give workers, users, suppliers, and local capital durable claims on the value they create.
Replicate the model through interoperable local firms rather than acquiring every successful node.
A company many families can own
Picture a neighborhood energy cooperative, a family-owned fabrication shop, a regional care network, or a software platform owned by the workers and businesses that use it. The legal wrapper could be a cooperative, employee ownership trust, benefit corporation, mutual, or a hybrid with carefully limited investor rights. The essential design is not the label. It is that the people exposed to the decisions have standing in them.
Shared upside
Profit-sharing, member dividends, or an ownership trust let productivity gains travel to the people who make the gains possible instead of stopping at the cap table.
Portable infrastructure
Open standards, exportable data, and multiple vendors make switching normal. A platform is useful because it coordinates people, not because it imprisons them.
Local judgment
Federated teams can adapt pricing, hiring, and service to their communities while sharing a common technical backbone and audited safety rules.
Patient capital
Retained earnings, community finance, and mission-locked capital trade some speed for resilience and reduce the pressure to become an acquisition target.
This is not romanticism about consensus. Democratic firms still need clear authority, competent operators, budgets, and mechanisms for resolving conflict. The hard question is how much coordination a network can sustain before it recreates a center. Platform-cooperativism research describes both the promise of shared ownership and the practical obstacles: participation is uneven, governance costs time, and incumbents have enormous capital and network advantages. Cooperatives can also struggle to raise risk capital without surrendering control, and member votes can slow a necessary pivot. Those are not footnotes; they are the design problems the alternative must solve with delegated authority, transparent mandates, durable reserves, and interoperable exit routes.
The morning after MANGO
In the thought experiment, there is no single replacement for the five firms. There are thousands of interoperable services: municipal and cooperative compute pools; independent model labs; chip designers licensed across manufacturers; local businesses that own the customer relationship; and research commons that make basic results auditable. People choose among providers the way they choose banks, farms, repair shops, and schools—with switching costs, yes, but without one universal gatekeeper.
Daily life would not become frictionless. Some products would be worse. There would be duplication, failed experiments, uneven quality, and political fights over standards. But failure would be more contained. A bad local provider would not automatically become a crisis for everyone who relies on one global dependency. Communities could fork a service, buy its assets, or build a competitor because the interfaces and skills remained available.
A practical startup blueprint
- Start with a real bottleneck. Solve a problem for a defined group of workers, households, or small businesses before optimizing for maximum addressable market.
- Separate the core from the moat. Keep safety-critical and coordination primitives open or portable. Compete on service, craft, trust, and execution.
- Write ownership into the first operating agreement. Do not promise that redistribution will happen after the exit; specify who earns votes, dividends, and data rights now.
- Scale by replication. Package training, software, procurement, and quality controls so another family or local team can launch without surrendering autonomy.
- Make concentration legible. Track dependency on any vendor, lender, cloud, model, or founder. Set thresholds that trigger diversification before a crisis.
- Use exponential tools for human-scale ends. AI and automation should increase the number of capable owners and reduce drudgery, not merely reduce headcount while increasing rent extraction.
What the research actually supports
The research does not prove that every cooperative will outperform every corporation, or that decentralization automatically produces innovation. It supports a narrower and more useful claim: market concentration and declining business dynamism are real policy concerns; shared ownership can improve workplace outcomes when paired with participation and good management; and platform cooperatives are a serious, imperfect alternative worth testing rather than a slogan to repeat.
That is enough to change the startup question. Instead of asking, “How do we become the one company that owns the category?” ask, “How do we become so useful that a thousand capable owners can build on the same pattern?” That is exponential growth without making everyone else’s agency asymptotically approach zero.
Sources and further reading
- OECD: Competition and market dynamism — concentration, markups, market-share instability, and business dynamism.
- OECD (2025), Concentration and business dynamics in product markets — product-market evidence across Europe and the United States.
- Autor, Dorn, Katz, Patterson & Van Reenen (2020), The Fall of the Labor Share and the Rise of Superstar Firms — Quarterly Journal of Economics.
- Bessen (2020), Industry Concentration and Information Technology — Journal of Law and Economics.
- Blasi, Freeman, Mackin & Kruse (2008), Creating a Bigger Pie? — NBER Working Paper 14230.
- ILO (2010), The Resilience of Social and Solidarity Enterprises — cooperative principles and resilience.
- Ghirlanda & Kirov (2024), Platform cooperativism through the lens of stakeholder theory — systematic literature review.
- DOJ and FTC (2023), Merger Guidelines — the agencies’ framework for concentration and competitive harm.
- The embedded talk on YouTube — starting at 20:07, as supplied for this post.