The Age of Extraction

1. Platforms: From Catalysts to Extractors

The power of the tech platforms relies on ancient economics: there have been essential platforms in every civilization.

Ancient origins. Platforms, defined as spaces or structures that bring groups together to transact or interact while reducing costs, are as old as civilization itself. From ancient Greek agoras to Middle Eastern bazaars, these "catalytic spaces" solved fundamental economic problems like matching buyers and sellers, building trust, reducing the minimum scale for small businesses, and fostering innovation. Their inherent neutrality allowed diverse activities to flourish, making them foundations of prosperous societies.

Modern evolution. The early internet and web embodied this spirit, promising a decentralized future where "small is the new big" and everyone could participate. However, modern tech platforms, unlike their historical counterparts, are private, profit-seeking corporations with grand ambitions. They have evolved from mere catalysts into sophisticated tools for extracting wealth and resources from the broader economy.

The shift to extraction. This transformation involves refining methods to take as much as possible—data, attention, and profit margins—from dependent businesses and users. This shift risks creating an economy perpetually unfair for much of humanity, dividing industries into extractors and their dependent agents, consumers, and employees.

2. Openness and Antitrust Fueled Early Tech Growth

The birth of the American software industry is a story of a push and pull between private industry and public government—and a story of unusual boldness and courage on both sides.

Platformization's genesis. The explosive growth of the American software industry in the late 1970s and early 1980s was not accidental but a direct result of "platformization." This involved creating open standards and operating systems that allowed thousands of independent software companies to emerge, writing "to the platform" rather than building their own hardware. This transformed computing from a niche activity into a multitrillion-dollar industry.

Antitrust as industrial policy. This openness was largely driven by aggressive government antitrust actions against dominant monopolies like IBM and AT&T.

• IBM was forced to "unbundle" its software from hardware in 1969, creating a market for independent software.
• AT&T, the telephone monopolist, was compelled by "net neutrality" rules in the 1970s to allow other businesses to operate "on top of" its network, leading to the rise of online services and the popular internet.
  These interventions, unlike the "national champion" policies in Europe and Japan, fostered competition and innovation, proving that restraining monopolistic power could spur economic growth.

A different path. While other nations focused on supporting their large national tech companies, the U.S. government's approach of breaking up monopolies and mandating open access created a fertile ground for new entrants. This "gardener" approach, pruning the largest trees to let others grow, led to American technological supremacy and an age of optimism about the internet's potential.

3. Optimism Faded: Platforms Became Monopolies

Not very often are so confident a set of predictions so wrong.

The illusion of decentralization. The early 2000s were marked by immense optimism about the internet's decentralizing power, with pundits proclaiming "small is the new big" and envisioning a world where individuals and small businesses would triumph over corporate giants. This belief was rooted in the open nature of the early web and the success of community-driven projects like Wikipedia.

The corporate pivot. However, this optimism failed to account for the inherent drive of private, profit-seeking corporations. Firms like Google, initially espousing "Don't be evil" and a user-first philosophy, eventually succumbed to shareholder pressures. The mission and money began to collide, leading to a strategic shift towards maximizing platform power and revenue.

Extraction through dominance. This shift was exemplified by Amazon's Marketplace, which initially enabled small sellers but then steadily increased fees, mandated advertising, cloned successful products, and harvested seller data. Similarly, social media platforms turned users into "unpaid or underpaid servants," harvesting their attention and data for advertising. The era of "blitzscaling" saw platforms aggressively acquire rivals and entrench their dominance, using scale as a weapon to "break free of competition entirely."

4. The "Everything Cocoon": Dependence by Design

You own a population first, then harvest it later.

Beyond traditional business. Modern tech platforms have transcended traditional business models, aspiring to become indispensable "everything cocoons" for as much of humanity as possible. This strategy, seen in Amazon's expansion into groceries and healthcare, or Apple's vast ecosystem, aims to create deep user allegiance and dependence, making switching platforms an "inconceivably exhausting undertaking."

The casino model. This approach is akin to a casino's poker room, where the goal is to get players to show up and stick around, regardless of individual wins or losses. The platform, like the casino, always takes its "vig" through subscription fees, advertising exposure, and data collection. The more time users spend on the platform, the more valuable they become as a raw resource to be mined.

The power of convenience. This dependence is largely built on human laziness and the irresistible force of convenience. Platforms make life so easy and comfortable that users become "couch-locked," unable to move to alternatives. While convenience offers personal liberation, it also creates an invisible source of power, actively shaping individual decisions and making us reliant on technological "armor" that, paradoxically, weakens our ability to live without it.

5. AI: A New Frontier for Platform Dominance

The big one is this: Who gets to be in the market for AI? What does it take to compete?

AI's dual origins. Artificial intelligence research has historically followed two paths: Symbolic AI, which codes explicit rules for problem-solving, and Connectionist AI (neural networks), which builds artificial brains that learn from data. The latter, initially discredited, resurfaced with "deep learning" in the 2010s, leading to breakthroughs like AlexNet and large language models (LLMs) such as ChatGPT.

Data and scale are paramount. Modern AI development relies heavily on two critical resources: gigantic, high-quality, pattern-rich datasets (like ImageNet or the vast language data for LLMs) and access to computing resources at scale (high-performance GPUs). This resource intensity raises the crucial question of who can compete in the AI market.

Reinforcing existing power. The major tech platforms, recognizing AI as a potential successor technology, have heavily invested in acquiring talent, data, and AI firms (e.g., Google buying Hinton's company and DeepMind, Microsoft's association with OpenAI). This "Kronos effect" aims to control and co-opt the AI challenge, ensuring it complements their core business models rather than displacing them. The risk is that AI will further entrench existing monopolies and marginalize human workers, turning us into "overworked middle managers" or creating emotional dependence.

6. Platformization Spreads Beyond Tech

But it would be a mistake to limit our understanding of the platform business model to the tech industries.

Beyond Silicon Valley. The platform model, perfected in the tech sector, has begun to spread to other, traditionally fragmented industries, signaling a broader economic transformation. This expansion centralizes economic power in sectors previously resistant to such consolidation.

Healthcare's transformation. In the U.S. healthcare industry, private equity firms have adopted a "platform" strategy, buying up and combining physician practices (e.g., anesthesiology, dermatology) into large groups.

• They promise doctors less administrative hassle and higher fees.
• In reality, these platforms raise prices for patients and insurers, prey on "out of network" billing, and impose grueling work requirements on doctors, leading to a loss of autonomy and quality of care.
• Doctors are often bound by noncompete clauses, effectively becoming a "laboring class" dependent on the platform.

Housing's centralization. Similarly, the single-family rental housing market, traditionally "mom & pop," has seen platformization by private equity firms like Invitation Homes.

• These firms bought thousands of foreclosed homes at deep discounts after the 2008 crisis.
• They use technology to manage properties at scale, systematically raising rents and implementing "junk fees" (e.g., "Lease Easy Bundle," "smart home" fees).
• They also cut costs by minimizing repairs, leading to tenant complaints and a transfer of wealth from renters to platform owners. This trend centralizes property ownership, eroding a core tenet of American economic identity.

7. Economic Imbalance: The Road to Authoritarianism

We are conducting a reckless economic experiment that history suggests has rarely gone well.

The fragility of inequality. History repeatedly demonstrates that societies with extreme economic imbalances are inherently fragile and prone to catastrophic collapse or the rise of authoritarianism. The pursuit of wealth and growth by tolerating such imbalances carries immense risk.

Lessons from Saint-Domingue. The 18th-century French colony of Saint-Domingue (Haiti) serves as an extreme cautionary tale. Despite being the wealthiest place in the Western Hemisphere due to highly centralized, scaled sugar and coffee plantations and brutal slave labor, its profound inequality led to a violent slave revolt, civil war, and ultimately, the devastation of the island. This illustrates how concentrated economies, while potentially productive in the short term, harbor inherent fragilities.

The sequence to serfdom. The "real road to serfdom" is a five-step sequence:

1. Monopolization: Unrestrained markets lead to dominant firms.
2. Extraction: Market power divides the economy into a narrow class of winners and a broad class of losers (consumers, workers, small businesses).
3. Mass Resentment: Systemic extraction fuels widespread anger and grievance, often directed at scapegoats.
4. Democratic Failure: Elected governments are unable or unwilling to address this resentment, often due to influence from the enriched class.
5. Rise of the Strongman: An authoritarian leader emerges, promising to address popular grievances, but ultimately centralizing power further.

8. History's Lesson: Balance Over Self-Correction

But history also tells us that there is a cycle and that over-centralization usually leads in two directions: a crash or long-term stagnation.

The myth of self-correction. For over a century, some economists have clung to the belief that markets are self-correcting, automatically dissipating monopolies and returning to a healthy equilibrium. Theories like "perfect competition," Galbraith's "countervailing power," and Kuznets's curve on inequality all posited that problems would naturally resolve "in the long run."

The "passive monopolist fallacy." These theories consistently fail because they ignore the active, predictable behavior of those who hold economic power. Monopolists are not passive; they vigorously defend their market position by:

• Buying out dangerous competitors.
• Controlling access to essential business resources.
• Diminishing labor unions.
• Enlisting government assistance.
  This active defense prevents market forces from naturally correcting imbalances, leading to persistent monopolies and widening inequality.

Deliberate action is required. History shows that economic imbalances do not disappear on their own. Instead, they require deliberate, structural interventions to rebalance power. Examples like Denmark's land reforms, post-WWII Germany's de-cartelization, and Taiwan's promotion of small businesses demonstrate that societies can achieve broad prosperity and equality through conscious policy choices, rather than waiting for an infallible market or state.

9. Tech Alone Won't Fix Inequality

History, unfortunately, gives us a more realistic picture of the impact of new technologies on the distribution of wealth in society.

The allure of technological fixes. In the face of growing economic inequality, there's a persistent hope that new technologies like artificial intelligence or cryptocurrencies might offer a solution. Sam Altman, for instance, vaguely asserts that generalized AI will eliminate global poverty, while projects like Worldcoin aim to use biometrics and blockchain to create a "more human economic system" and provide universal access to the global economy.

AI's ambiguous promise. While AI can create wealth, its impact on wealth distribution is historically mixed. Like the agricultural plow, some technologies spread productivity widely. Others, like the cotton gin, exacerbate existing inequalities. AI could augment human workers, making them more productive, or it could replace them en masse, leading to widespread marginalization. The outcome depends heavily on how AI is designed and where economic power resides.

Crypto's centralization challenge. Cryptocurrencies, particularly Bitcoin, were conceived as a decentralized counterweight to traditional financial institutions, aiming to reduce political manipulation and favoritism. However, the crypto movement has struggled with its own centralization, exemplified by the collapse of exchanges like Mt. Gox and FTX, where centralized control led to massive fraud and theft. While crypto has created wealth for some, it has yet to make a significant dent in global wealth imbalance or provide accessible solutions for the truly poor.

10. Rebuilding Equality: Anti-Monopoly and Neutrality

A government devoted to the long-term prosperity of its people must have in place structures that are committed to the constant balancing of the economy—the equivalent to checks and balances for political power.

The gardener state. The ideal economic state is not a "night watchman" that passively observes, but a "gardener" that actively prunes overgrown plants (monopolies) to ensure sunlight reaches all others. This involves structural balancing, akin to political checks and balances, to prevent the aggregation of economic power.

Anti-monopoly's resurgence. A strong and ongoing antitrust enforcement program is crucial. Claims of "natural monopoly" in tech are often premature and self-fulfilling prophecies, as demonstrated by the rise of competitors to eBay or the potential for alternatives to Google Search. Antitrust doesn't create competition but removes the tools for killing it, forcing dominant firms to compete on merit and creating openings for industrial succession. The recent antitrust actions against Google, Amazon, and Facebook signal a welcome return to this vital policy.

Neutrality rules for public callings. Essential platforms, like historical inns or blacksmiths, should be recognized as "public callings" with duties to serve all paying customers without arbitrary discrimination. Net neutrality rules, which were vital for the early internet, exemplify this principle by preventing essential infrastructure from picking and choosing whose traffic to carry. Applying similar neutrality rules to modern tech platforms would prevent them from distorting competition and extracting undue value by favoring their own products or partners.

11. Empowering Countervailing Forces and Capping Extraction

The power of a giant firm like General Motors of the 1950s was constrained not just by its direct competitors (like Ford) but also by its suppliers.

Revisiting countervailing power. John Kenneth Galbraith's concept of "countervailing power" highlights that the economic might of large corporations can be constrained not just by direct competitors, but also by those they bargain with, such as suppliers and labor. While Galbraith mistakenly believed this power would arise automatically, it remains an essential element in restraining platform power.

Empowering key stakeholders. To rebalance the platform economy, the law should actively empower those in a position to constrain platforms:

• Owners of valuable information: Copyright holders (e.g., news organizations, artists, authors) whose content is essential for training AI models can collectively bargain to limit platform extraction.
• Workers: Organized labor (unions) can provide a crucial check on employer power, ensuring fair wages and working conditions. The decline in unionization has coincided with stagnant wages and rising inequality, underscoring the need for stronger employee representation, even for professionals.

Utility rules and quarantines. When market remedies fail, essential platforms may need to be declared utilities, with hard caps on their extraction.

• Price caps: Europe's cap on credit card transaction fees (0.3%) provides a model for limiting hidden taxes imposed by platforms that operate like "toll bridges."
• Line of business restrictions: Quarantines can prevent platforms from leveraging their monopoly in one area to dominate another. For example, banning Amazon from selling its own products in direct competition with its Marketplace sellers, or preventing existing tech giants from acquiring or controlling new AI platforms, would foster independent innovation and prevent further centralization.

Last updated: February 5, 2026