The propagation principle has a sharp consequence the moment it meets economics, because the industrial business model is built on almost the opposite instinct. The classic firm concentrates capacity inside itself and monetizes controlled access to it; the model asks a different question of the same activity:
Does the business merely accumulate capability, or does it increase the capability of the wider ecology?
This does not turn businesses into charities, and it does not ask anyone to give everything away. Profit remains, ownership remains, legitimate intellectual property remains, aggressive competition remains. What changes is the definition of a successful firm — and once that shifts, a great deal follows.
The central question shifts
The industrial firm runs, roughly, on a chain: gather capital, acquire labor, control tools, concentrate knowledge, produce output, protect advantage, retain surplus, scale control. Its competitive advantage usually rests on some asymmetry — proprietary knowledge, exclusive distribution, patents, locked-in customers, switching costs, scarce infrastructure — and the stronger the moat, the safer the business. From the firm’s own perspective that is entirely rational. But from an ecological perspective a very successful firm can be a capacity-concentrating engine: the company grows steadily more capable while its customers, workers, and suppliers grow steadily more dependent.
So the model keeps the industrial question — how much value did we capture? — and adds a second one beside it: how much new capacity did our activity leave distributed throughout the ecology? A firm can still earn, own, and compete; but its ecological contribution turns on whether the relationships it creates leave others more capable or more dependent. That single distinction reorganizes everything below.
Customer as participant, worker left more capable
Industrial logic treats the customer as the end of the chain — company produces, customer consumes. The model prefers, where it applies, company produces, customer gains capability, new activity becomes possible. Two software products can both be profitable while one makes the customer ever more dependent on the vendor for each new function and the other expands what the customer can independently do; the second leaves more distributed capacity behind it. Not every product must teach its user to reproduce it — a restaurant need not turn diners into chefs — but the broader question always applies: what becomes possible in the customer’s ecology because this relationship exists?
The worker relationship changes even more. A firm buys a person’s useful capacity during employment, and when they leave, much of what they gained may be nonportable, confidential, tied to internal systems, or legible only through the firm’s own status. The model asks a question ordinary accounting never poses: did the participant leave more capable than they arrived? — with more skill, judgment, relationships, reputation, and future optionality. A company that consumes twenty years of a life and leaves a person narrowly adapted to one internal system may post excellent returns while depleting human ecological capacity, which the model counts as a liability. This is why development stops being merely a cost: training produces portable ecological capacity, so the fearful question what if we train them and they leave? meets its inversion — what if we don’t train them and they stay? — and, deeper still, what happens to the whole ecology if every organization tries to capture development rather than circulate it?
Proportionate exclusivity
The industrial firm often benefits from preventing useful knowledge from spreading, and sometimes legitimately so — without some protection, the investment that creates capacity may never be made. The model does not require abolishing intellectual property; it asks whether the degree and duration of exclusivity stay proportionate to the function they serve:
Temporary concentration may be justified when necessary to create capacity; permanent concentration requires much stronger justification.
That is a different default from maximum exclusion is naturally desirable. A firm may need enough exclusivity to recover its investment, after which the wider ecology tends to benefit from licensing, standards, interoperability, publication, training, open interfaces, and diffusion — the same friction-proportionality logic applied to knowledge: no more restriction than the consequence actually requires.
Two kinds of moat
This is one of the largest departures. Industrial strategy says build a moat; make it hard for others to compete. The model does not reject defensibility, but it distinguishes two kinds. A healthy moat comes from excellence, trust, accumulated skill, quality, reputation, and continuous innovation — it says stay because we’re good. A problematic moat comes from trapping users, controlling necessary infrastructure, suppressing interoperability, buying competitors, monopolizing essential signals, and manufacturing switching costs — it says stay because leaving is difficult. The model strongly prefers the first, which makes lock-in an ecological liability rather than an asset: it follows straight from traversability that lock-in reduces a participant’s ability to move through the ecology, so firms would be expected to support reasonable data and account portability, interoperability, cancellation, migration, and exit. The customer relationship becomes more legitimate precisely because staying becomes more voluntary.
Turned around, this yields a positive form the model actively favors — a generative moat. A business that increases ecological capacity develops capable users, capable suppliers, talented alumni, trusted communities, and interoperable systems, and those participants voluntarily return, collaborate, recommend, and build around it. The alternative to lock-in is not weakness; it is becoming such a useful node that others repeatedly choose to connect. Instead of “you cannot leave me,” it says “you can leave, and this remains a valuable place to return to.”
Platforms and AI firms at the propagation layer
Some businesses sit exactly at the layer this function governs. Where industrial firms controlled goods, platforms control connections — who gets discovered, who gets connected, who receives attention, which creators survive, which communities form, which businesses reach customers. That is not merely private product design; it is ecological routing, so a platform’s success can no longer be read off engagement, revenue, and ad yield alone. The model also asks whether it distributes opportunity, creates new capable nodes, over-concentrates attention, lets participants leave with their network, and leaves creators dependent on opaque rules — a challenge to much of today’s platform economy.
The AI firm faces the deepest version. Industrial logic says maximize proprietary advantage; the model asks how much of that machine-generated capacity actually enters the wider ecology? If AI lets one company replace a hundred thousand workers and concentrate all the gains internally, productive capacity soars while ecological capacity may fall; if the same technology gives millions of participants real access to knowledge, design, education, creation, and research, machine capacity is becoming distributed ecological capacity. That distinction — the same distribution question the connection function ends on — may be central to post-AI economics, and it is why ownership is no longer the only relationship that matters: as a firm’s capability grows in consequence (and especially as its technology becomes essential infrastructure), the question becomes what standing does ownership legitimately confer, and what consequences create standing for others?
Growth, competition, and failure
Several familiar business concepts shift in meaning. Growth industrially means more revenue, share, customers, assets, valuation; the model adds did the business increase the ecology’s total capacity? A firm can grow financially while shrinking ecological capacity — by destroying competitors, capturing infrastructure, deskilling workers, locking customers in, exhausting resources, making itself indispensable — and a relatively small firm can produce enormous ecological growth by creating tools, standards, skills, networks, and new independent businesses. That changes what scale means, and it lets the model recognize ecosystem creation as a major form of success: the exceptional firm may even create its own competitors, as internet infrastructure did, generating far more total capability than any one vertically integrated owner could. Competition remains, and the model can be strongly pro-competitive — competition exposes poor performance, drives innovation, prevents dependency, and often reduces concentration — but its purpose shifts from destroying rivals toward being one mechanism through which the ecology discovers better forms.
Failure changes most of all. Industrial entrepreneurship is throttled by personal risk: lose the business and you may lose income, healthcare, housing, and status. Because baseline standing separates entrepreneurial failure from personal ecological exile, people can attempt businesses without annihilation on the downside — so the model could make markets more dynamic and experimental while simultaneously reducing their power to determine anyone’s basic standing. The model preserves standing and reentry pathways, not every firm.
The firm as a temporary concentration
Underneath all of this is a quiet reconceptualization. Rather than a quasi-permanent entity whose natural goal is endless self-expansion, the model can see a firm as a temporary concentration of resources, people, technology, and jurisdiction assembled to perform particular functions. The governing question becomes is this concentration still useful? — if so, continue; if the function is complete or better forms emerge, the resources can move elsewhere. Corporations become less ontologically sacred: ecological structures, not organisms entitled to survival at all costs.
That reframing bounds the firm’s claims on people. The corporation does not own its participants — an employee brings capacity to the project and the company pays for agreed participation, but it acquires no ownership of a person’s identity, political belief, entire future, unused creativity, relationships, or interior reality. Compensation, too, splits into what industrial wages fused: today a wage both rewards contribution and supplies access to basic life, and the model separates them — basic standing supplies baseline viability, so pay can attach more honestly to contribution, skill, scarcity, responsibility, risk, and value created. The employer stops implicitly saying obey because I control your access to survival, and negotiation becomes less coercive. In a post-work ecology this reverses power further: when humans do not need every available job, organizations wanting excellent human participation must compete for it with interesting work, real responsibility, development, autonomy, recognition, and good collaborators — which could profoundly improve organizational design.
A richer balance sheet, and the wake
All of this points toward a fuller account of what a firm produced. Alongside financial return (did it stay viable?) and productive return (what useful goods or services were created?), an ecological balance sheet would track participant capacity (did employees, users, and partners become more capable?), network effects (did useful relationships and institutions emerge?), renewability (did it regenerate or deplete what it depended on?), propagation (did capacity spread beyond the firm?), dependency (how much unnecessary reliance did it create?), and future optionality (did it open or close future pathways?). Profit stays a vital local signal — it shows the firm is generating enough valued exchange to cover its costs under current conditions — but it cannot show whether workers were depleted, externalities exported, users trapped, future capacity destroyed, or public infrastructure quietly carrying the costs. Profit remains a vital local metric, not a complete ecological one.
Two consequences complete the picture. Shareholder primacy becomes hard to defend once the firm is seen inside its full web of relationships — owners, workers, customers, suppliers, communities, public infrastructure, future participants. Shareholders hold real standing because they provide capital and bear financial risk, but not exclusive standing; and rather than vague stakeholder rhetoric, the model favors precise multi-standing governance: which groups have standing over which decisions — shareholders over some capital questions, workers over certain workplace ones, customers over data and product safety, communities where externalities arise. And the firm no longer gets to define efficiency alone. Industrial accounting lets a company announce it became thirty percent more efficient; the model asks where did the removed cost go? — was work shifted onto customers, emissions displaced elsewhere, social costs pushed onto the public, caregiving burdens created by outsourcing? Efficiency has to be traced across the ecology, because a local optimum can be an ecological loss. The firm becomes responsible for its wake — not that every downstream effect is legally chargeable to it (that would make enterprise impossible), but that consequential externalities are not imaginary merely because the ledger ignores them, which makes good ecological accounting a matter of accurate systems intelligence rather than virtue.
From extraction to propagation
Compressed to its arc, the shift is clear. The industrial model runs: concentrate resources, organize labor, produce output, capture value, defend concentration, expand. The model-compatible, AI-era firm runs: assemble capacity, create useful output, develop participants, propagate capability, retain enough value for renewal, and remain valuable enough that the network keeps choosing the relationship. Profit, ownership, competition, and innovation all survive the translation; what changes is that the ecology around the business becomes structural rather than an afterthought. And the deepest difference is a reversal of the definition of success itself:
The industrial company often succeeds by making itself increasingly necessary. The highest form of the model’s firm succeeds by making the ecology increasingly capable — even where that means participants grow progressively less dependent on the company itself.
That is a very different theory of enterprise, and in a world where AI and robotics make productive capability vastly easier to generate, it may turn out to be far more realistic than it first sounds.