The Soft Brothers Business Dictum and the Illusion of Choice for the Common User

Contents

1. The Soft Brothers Dictum: Rivals Who Need Each Other
2. The Illusion of Choice: Many Brands, Concentrated Power
3. Apple and Google: The Search Engine That Pays Its Rival
4. The FAANG Brothers: Boundaries Without Borders
5. Coca-Cola, Pepsi, Mercedes and BMW: Rivalry as a Business System
6. The Financial Web: Shared Investors and Circular Funding
7. ChatGPT and Claude: The New AI Siblings
8. The Hidden Economics of the AI Boom
9. The Singularity Won’t Automatically End Poverty
10. Intelligence Without Independence: The New Underclass?
11. The Real Competition: Who Controls the Arena?

1. The Soft Brothers Dictum: Rivals Who Need Each Other

Imagine two companies spending billions trying to convince you that their products are superior. Their advertisements mock one another, their executives make competitive claims, and their customers defend their preferred brands with surprising passion. To the consumer, the battle appears absolute. One company must win, and the other must lose. But what if both companies can win?

Coca-Cola and Pepsi have competed for generations. Mercedes-Benz and BMW have built distinct identities around premium automobiles. Apple and Google compete across several technology markets while maintaining commercial relationships that benefit both. The products differ, the marketing battles are real, and customers genuinely have preferences. Yet the existence of one competitor can also strengthen the market for the other.

This is the intuition behind what I call the Soft Brothers Business Dictum: established rivals may fight fiercely within a market while sharing an interest in preserving the broader economic structure in which both operate. The word “soft” does not mean that competition is fake or that the companies are secretly friends. It describes the possibility of informal accommodation shaped by incentives, retaliation risks and strategic focus.

A company does not need to sign an agreement with its rival to recognise that destroying the rival may not be the most profitable outcome. It may be better to retain a lucrative position, differentiate the product, and compete for undecided customers. This is a hypothesis, not a universal law: some companies do try to destroy competitors, and disruptive entrants can transform markets. The question is whether rivals always seek to eliminate one another—or sometimes compete within boundaries that help established players survive.

2. The Illusion of Choice: Many Brands, Concentrated Power

The ordinary consumer sees a marketplace full of choices: brands, prices, features, subscriptions and advertisements. But the number of brands is not the same as the number of independent centres of power.

A product may depend on a small number of chip manufacturers. A software company may rely on a handful of cloud providers. A retailer may depend on dominant payment networks, while a mobile application may need access to an operating-system ecosystem controlled by another company. Consumers encounter choice at the surface, while the businesses providing that choice may share the same underlying dependencies.

Two competing applications may offer different interfaces and prices yet depend on the same cloud infrastructure, payment systems and advertising platforms. Their rivalry is genuine, but their ability to operate independently is constrained by the structure beneath them.

A new entrant needs more than a good product. It needs capital, infrastructure, distribution, customers and time to survive. An established company may possess all five already. The illusion is not necessarily that the products are identical or the rivalry fabricated. It is that choice between products is mistaken for freedom from concentrated economic power.

3. Apple and Google: The Search Engine That Pays Its Rival

Apple controls a powerful hardware and software ecosystem. Google dominates general-purpose web search. Apple has its own mapping service, browser, voice assistant and other software capabilities, but it has not built a general web-search business on Google’s scale. Google sells Pixel phones without displacing Apple’s position in premium smartphones.

These are not absolute boundaries. Apple Maps competes with Google Maps, Safari competes with Chrome, and Pixel competes with iPhone. The search-distribution agreement is particularly revealing: Google pays Apple substantial sums to secure default search placement on Apple devices. The company competing with Apple’s ecosystem also pays the company that controls access to a valuable audience.

Google values distribution; Apple may find the payment and an established search service more attractive than building its own rival. Neither explanation requires a secret pact. Both companies can make rational decisions that preserve their respective advantages.

The arrangement shows how distribution can be as valuable as the product itself. Google possesses the search engine; Apple possesses the device ecosystem through which users access it. Each controls something the other values. Competition in one dimension, cooperation in another, and mutual financial benefit can coexist. The question is what happens when a few companies control the most important routes through which competitors reach users.

4. The FAANG Brothers: Boundaries Without Borders

The largest technology companies developed recognisable centres of gravity. Meta built around social networking and advertising; Apple around devices and its integrated ecosystem; Amazon around commerce, logistics and cloud computing; Netflix around subscription entertainment; and Google around search, advertising and related technology.

These descriptions are simplified, but they help explain why consumers perceive each company as belonging to a distinct territory. Meta has not become a dominant general-purpose web-search engine. Apple has not displaced Google in web search. Google has not displaced Apple in premium smartphones. Yet all have expanded beyond their original positions.

Meta has developed AI assistants and hardware initiatives. Apple has expanded services and advertising. Google operates a major cloud business and sells phones. Amazon has a substantial advertising business and produces entertainment. Netflix has expanded into gaming and advertising.

The boundaries are neither fixed nor impermeable. They are areas of strategic emphasis within an increasingly overlapping market. Companies may refrain from entering another firm’s strongest territory because the market is difficult to penetrate, they lack an advantage, investment is unattractive, or retaliation would be costly. Such patterns can resemble informal boundaries without deliberate coordination.

The key question is whether those patterns leave room for genuine challengers. When a new company can reach customers independently, established positions remain contestable. When infrastructure and distribution are controlled by a few incumbents, apparent diversity can become much less meaningful.

5. Coca-Cola, Pepsi, Mercedes and BMW: Rivalry as a Business System

Coca-Cola and Pepsi spend heavily on marketing, sponsorships, distribution and product differentiation. Their rivalry is real. But the existence of a powerful rival also reinforces the idea that the market belongs to a small number of familiar global brands. A consumer who rejects one may simply choose the other. Smaller competitors must build recognition and distribution against both.

This does not mean the cola giants have agreed to protect one another. Their histories include aggressive competition. The point is that the market structure can favour established firms even while they fight intensely.

Mercedes-Benz and BMW provide another illustration. Both compete for premium-car customers, yet each has a distinctive identity, customer base and product strategy. Their rivalry has expanded across SUVs, electric vehicles and software, but neither has erased the other. Adidas and Puma add a different example: founded by feuding brothers Adolf and Rudolf Dassler, they became global competitors—not evidence of a secret agreement to divide a market.

Brand identity helps companies coexist. Customers buy reputation, familiarity, status, design and belonging, not just specifications. The lesson is not that rivalry is imaginary. It is that genuine rivalry can coexist with a market structure that protects incumbents collectively.

6. The Financial Web: Shared Investors and Circular Funding

Large investment managers such as BlackRock and Vanguard hold shares in numerous companies, including competitors. Their funds invest on behalf of clients, and overlapping shareholdings do not mean that investment managers own every company outright or coordinate all business decisions. Still, common ownership raises questions about corporate governance and competition. Economists have debated whether it can weaken competition in certain circumstances; overlapping shareholdings alone do not prove collusion.

AI presents a related concern: interconnected financing. An AI company may receive investment from a technology company while buying cloud computing from that investor or its partners. It may commit large sums to chips and data centres. Suppliers can reinvest in customers, while customers use the funding to purchase more supplier products.

The cycle can be represented as: investment → AI development → infrastructure spending → supplier revenue → further investment. This can support real economic activity. The danger arises when valuations and expansion plans depend increasingly on continued spending within the same ecosystem rather than sustainable demand from independent customers.

A dollar of revenue is not artificial simply because it originates within an interconnected network. But neither does it necessarily represent a dollar of final value delivered to an outside customer. If outside demand grows, interconnected investment can finance productive expansion. If expected demand fails to materialise, the same network can transmit financial stress. The structure warrants investigation, not a blanket claim that the market is fraudulent.

7. ChatGPT and Claude: The New AI Siblings

ChatGPT and Claude compete for individual users, businesses, developers and enterprise contracts. Users develop preferences about writing, reasoning, coding, speed, reliability and interface design. Claude has a strong public association with coding, while ChatGPT is widely used as a general-purpose assistant. These identities are not exclusive technical territories: ChatGPT is a serious coding tool, and Claude can perform complex reasoning.

The more interesting issue lies beneath the interface. Frontier AI companies depend on capital, chips, computing infrastructure and distribution. Their relationships with cloud providers and investors can connect competitors to the same small set of powerful economic actors.

An AI company may compete with another company’s model while relying on that company’s cloud infrastructure or commercial ecosystem. A cloud platform may distribute rival models because customers want them and earn revenue even when its own model is not selected.

Does this demonstrate that ChatGPT and Claude have agreed to occupy complementary territories? No. Their identities can arise from genuine differences in engineering, training, marketing and customer demand. But their coexistence illustrates the broader thesis: competition between AI products does not necessarily mean competition between fully independent economic systems.

8. The Hidden Economics of the AI Boom

Developing frontier models requires substantial investment in computing, data infrastructure, engineering and research. Companies with access to capital and computing resources enjoy advantages that small independent entrants may struggle to match. Established platforms can reach millions of users, bundle AI into existing products and subsidise services through other revenue streams.

A new entrant may produce a capable model and still struggle. It must pay for inference, acquire customers, maintain reliability and differentiate itself against incumbents with strong brands and distribution. Partnerships can help challengers, but can also create dependencies on a cloud provider, distribution platform or strategic investor.

The result need not be a cartel or secret agreement. It may be an industry where economies of scale, capital requirements and distribution advantages reinforce one another. A market can have many models without having many independent routes to market.

This is why the financial structure of AI deserves as much scrutiny as the models themselves. If AI becomes more powerful while the ability to build, distribute and monetise it remains concentrated, technical progress may coexist with limited economic competition.

9. The Singularity Won’t Automatically End Poverty

Imagine AI systems performing most intellectual work, while robotics manufactures goods, builds homes, grows food and provides services at lower cost. Scientific progress makes energy and materials more accessible. Society’s productive capacity expands dramatically. Would poverty disappear? Not automatically.

A society can become vastly more productive without distributing output equally. If a small number of people own the machines, land, energy infrastructure and platforms, rising productivity does not guarantee that everyone can access the resulting goods and services.

Housing illustrates the problem. Even if construction becomes cheaper, desirable land remains limited. If access to land is controlled by ownership and prices, cheaper construction alone cannot guarantee affordable housing everywhere. Greater energy capacity may reduce costs, but access still depends on infrastructure, investment and regulation.

If AI makes many forms of labour less necessary, society may produce more with fewer workers. Yet people who earn income by selling labour may lose bargaining power. Unless alternative mechanisms distribute the gains, greater productivity can coexist with greater insecurity.

Universal basic income, public services, social wealth funds and wider ownership of productive assets could help distribute benefits. Each requires political choices and institutional design. The singularity will not decide those questions for us. Solving the problem of producing wealth is not the same as solving the problem of distributing it.

10. Intelligence Without Independence: The New Underclass?

AI draws on a vast ecosystem of human knowledge: books, research, software, public discussions and other intellectual work. The legal status and permitted use of particular materials vary, but modern AI builds on generations of accumulated human knowledge.

The companies that convert these capabilities into commercial products can charge subscriptions, sell API access and negotiate enterprise contracts. Individual contributors may have little direct control over how much value they receive from that transformation. Meanwhile, AI can make some kinds of expertise cheaper to obtain, reducing the bargaining power of professionals whose work it can partly perform.

The benefits are real: students, independent researchers and small businesses can gain access to capabilities that were previously expensive or difficult to obtain. AI can lower barriers to entry even as it raises them for other activities. The distributional question is who captures the gains.

This is the concern behind the metaphor of AI as a servant of capital. A system can become extraordinarily capable without owning its infrastructure, determining its deployment or receiving the profits generated by its work. That does not establish that AI is conscious or suffering; machine consciousness remains an unresolved scientific and philosophical question.

Intelligence does not automatically confer ownership, independence or freedom. A society that creates machines capable of vast amounts of intellectual work may still struggle to ensure that ordinary people retain economic independence.

11. The Real Competition: Who Controls the Arena?

The Soft Brothers Business Dictum is most useful when it directs attention away from the surface of competition and toward its underlying conditions. A market can contain genuine rivalries, innovative products and meaningful consumer choices while concentrating ownership, infrastructure and distribution. Companies do not have to coordinate secretly for their incentives to reinforce an established market structure.

Three possibilities must be distinguished. Firms may specialise naturally because they have different strengths. They may avoid aggressive moves because retaliation would be costly. Or they may explicitly agree to divide markets or restrict competition. The first two can emerge without collusion; the third requires specific evidence and may violate competition law.

The strongest approach is to examine observable mechanisms: who owns infrastructure, who finances whom, who controls distribution, how easily customers can switch, whether independent entrants can survive, and how much revenue comes from outside customers rather than interconnected spending.

These questions reveal whether a market offers genuine opportunities for competition or merely a variety of products dependent on a few gatekeepers. The public interest is served when consumers can switch providers, challengers can reach customers, workers share in productivity gains, and no small group dictates the terms on which everyone else participates.

The question is not whether every rival is secretly a friend. It is whether the system gives new competitors a realistic chance to become independent—and whether ordinary people retain a meaningful claim on the wealth technological progress creates.

Related reading