The Power Paradox: Why Decentralization Doesn't Mean Giving Up Efficiency

By Vitalik Buterin

There’s a tension lurking at the heart of modern society that most people feel but rarely articulate: we want powerful systems to drive progress, yet we fear what happens when power concentrates too much in any single place. This isn’t just political philosophy—it’s the core challenge facing blockchain technology and decentralized systems.

The Three-Way Tension Nobody Talks About

Most of us harbor simultaneous fears about three different power centers. We appreciate what large corporations produce—their innovation, scale, and efficiency—yet we despise monopolistic control, manipulative business models, and the way mega-corporations distort entire markets and governments to serve their interests.

Similarly, we need governments to maintain order and provide public goods. But we cringe at their ability to restrict speech, control information, arbitrarily pick winners and losers, and abuse fundamental freedoms.

Then there’s the third threat: uncoordinated mass movements. Civil society organizations and independent institutions are valuable, but history shows us the dangers of mob rule, cultural purges, and populist movements that claim to represent “the people” while actually serving a single agenda.

The uncomfortable truth? All three forces—Big Business, Big Government, and collective mob action—have historically been engines of progress. Yet all three can become oppressive when unchecked.

Why Natural Monopolies Keep Getting Bigger

The 20th century rise of the United States and the 21st century ascent of China can be explained by a single economic principle: economies of scale. Countries and platforms that achieve scale earlier compound that advantage exponentially. If I control twice your resources today, I can generate more than double your progress by next year—meaning by the year after, I’ll have even more than twice your resources. Over time, this mathematical reality means whoever gets ahead first will eventually control everything.

This isn’t a bug in capitalism; it’s baked into the system. A corporation valued at $1 billion will outspend 100 smaller competitors ($10 million each) on market control, environmental shaping, and innovation bundling. This explains natural monopolies—sectors where scale advantages are so overwhelming that consolidation becomes inevitable: telecommunications, utilities, digital platforms.

Historically, two forces prevented complete monopolization:

Diseconomies of scale: Large institutions become bureaucratic, slow, and inefficient. Coordination costs rise. Internal politics increase. Geography matters. These friction costs slowed dominant players.

Diffusion effects: Ideas, talent, and technology leak across boundaries. Developers switch jobs and bring their knowledge elsewhere. Countries reverse-engineer successful products. Rival platforms cross-pollinate features. This acted like a rubber hand pulling laggards closer to leaders.

But in recent years, this equilibrium has shifted dramatically. Technology advancement has made economies of scale more powerful than ever. Automation eliminates coordination costs. Most critically, proprietary technologies (closed software and hardware ecosystems) prevent diffusion like nothing before. You can’t inspect what you can’t access. You can’t reverse-engineer what’s locked behind code you can’t see. Distribution no longer requires giving away control—Amazon doesn’t have to let you open their servers; Netflix doesn’t have to explain their algorithm.

The result: diffusion effects have weakened precisely when economies of scale have strengthened.

What Happens When Scale Goes Unchecked

When one entity gets too large, it inevitably distorts its environment to maximize profits at the expense of everyone else. A monopoly raises prices above marginal cost—extracting consumer surplus. A mega-corporation lobbies governments to bend rules in its favor. A platform manipulates culture through algorithmic curation.

This creates two specific problems:

The “Evil” Problem: As businesses grow, their profit motive increasingly diverges from user welfare. Early stage industries are driven by enthusiasts and genuine innovation. Later stage industries become extraction machines. Gaming went from “fun and achievement” to “psychological manipulation and slot machine mechanics.” Prediction markets went from “improving governance” to “sports betting platforms.” Crypto went from “empowering individuals” to—well, you know.

The “Soullessness” Problem: Homogenization. When multiple dominant players share the same profit motive and face identical competitive pressures, they converge to the same strategies. Urban architecture becomes identical Starbucks. Hollywood produces the same plot repeated 50 times. Platforms all adopt the same engagement metrics. This isn’t evil exactly—it’s institutional commonality, where reasonable actors in similar conditions make similar choices.

The Decentralization Solution Nobody Fully Grasps

The solution isn’t to eliminate large-scale systems—they drive real progress. The solution is to forcibly redistribute the means of production and control mechanisms so that scale advantages don’t concentrate power in fewer hands.

This can happen several ways:

Policy-level interventions: The EU’s mandatory USB-C standard makes it harder to build proprietary lock-in ecosystems. China’s technology transfer requirements force knowledge diffusion. The U.S. ban on non-compete agreements means employees can take their skills elsewhere, spreading tacit knowledge across the economy. Copyleft licenses (like GPL) require that improvements built on open-source code must remain open-source.

Novel mechanisms: Governments could tax products based on their “proprietary degree”—charging more for closed systems, zero for open-sourced ones. Intellectual property Harberger taxes could incentivize efficient knowledge utilization.

Adversarial interoperability: This is the key innovation. As Cory Doctorow describes it, adversarial interoperability means building products that interface with existing systems without permission. Third-party printer ink. Alternative app stores. Independent repair shops using compatible parts. Browser extensions that block AI-generated content on platforms. Decentralized stablecoin exchanges that bypass the “single point of failure” risk in centralized finance.

The genius of adversarial interoperability is that it operates at the interface layer—where most Web2 value extraction happens. Users can stay within networks they value while avoiding the platform’s rent-extraction mechanisms. It’s not replacing the system; it’s providing alternative ways to interact with it.

Diversity and cross-cutting collaboration: Glen Weyl and Audrey Tang’s concept of “facilitating collaboration between differences” helps large organizations (open-source communities, national alliances, international standards bodies) achieve economies of scale benefits while remaining competitive against centralized giants. The key is internal pluralism—multiple decision-making centers, different value systems, diverse goals working in parallel rather than unified.

The Blockchain Application

Ethereum’s Lido represents a case study in this framework. A single staking pool controlling 24% of network ETH could theoretically create centralization risk—but in practice, concerns are far lower because Lido isn’t a single entity. It’s a DAO with dozens of independent node operators, dual governance allowing ETH stakers veto power, and commitment to remaining below safe concentration levels. The structure itself prevents the power that comes with scale from concentrating into a single decision-making node.

This is the template: you can have large scale without power concentration if you deliberately design for distributed control from the ground up.

The Unresolved Challenge

There’s a theoretical threat called the “fragile world hypothesis”: as technology advances, more entities gain the capability to cause catastrophic harm. A highly fragmented world might choose to use those capabilities. Some argue the only answer is ultra-concentration of power—so one actor can prevent others from harming everyone.

But this gets the incentives backwards. The more concentrated power becomes, the more likely it gets used for harm, because the concentrated actor faces fewer checks and balances. The real insurance policy is defensive acceleration—building defensive technologies that remain open and accessible to everyone, developing in lockstep with offensive capabilities. This reduces the security anxiety that drives power concentration in the first place.

The Moral Framework

Rather than “you must become powerful” (master morality) or “you can never be powerful” (slave morality), the framework emerging from blockchain communities suggests: you should pursue positive impact and empower others, but you cannot form hegemony.

This is the distinction between empowerment rights (the ability to create value and grow) and control rights (the ability to exclude others and lock them in). You can have one without the other.

The challenge: How do you retain the flexibility and decisiveness advantages that come with concentrated authority while preventing the power concentration itself from becoming dangerous? For some sectors—like open protocols (TCP, IP, HTTP) or English as a lingua franca—decentralization happens naturally. For others—where intentional direction and coordinated action matter—it remains genuinely hard.

This isn’t a solved problem. But it’s the right problem to be solving.

The core insight: decentralization doesn’t sacrifice progress; it redistributes who captures the gains from progress, and who bears the risks.