Why Permissionless Infrastructure Wins

Ethereum is replaying the Internet and Linux.

“Stripe wants everything to happen on Tempo, but JP Morgan wants everything to happen on JP Morgan Chain, and Circle wants everything to happen on Arc, and so on and so forth. They will never agree. The large players will never agree to build on the infrastructure of another large player. This is why Ethereum is the only option. It’s the only way forward as the neutral infrastructure that everyone can agree on.”

– Alex Gluchowski, founder of zkSync

In 1995, much of the technology establishment was convinced the Internet would lose to proprietary corporate networks. They were wrong, and Ethereum’s critics today are likely to be wrong for similar reasons. The highest profile example was Bill Gates, who predicted in his book The Road Ahead that the future of digital commerce would not run on the open Internet but on proprietary networks owned by companies like Microsoft and Oracle. This was the consensus. As a16z co-founder Ben Horowitz writes, “almost nobody thought the Internet would be significant beyond the scientific community–least of all the most important technology industry leaders who were busy building proprietary alternatives.” The same thing happened with Linux. Through the late 1990s, Sun Microsystems dominated the market for high-end Unix servers, but by the early 2000s it had lost most of that business to open-source Linux running on cheap commodity hardware.

The same pattern is playing out in financial infrastructure today. Corporations, sensing opportunity and threat, race to build proprietary blockchains inside walls they control. For a while, the proprietary versions look like they’re winning – they’re faster, with a better UX, and have large business development teams. Then they are slowly eaten by an open, credibly-neutral alternative because no company can keep pace with permissionless innovation forever and no serious participant will build on infrastructure a competitor controls.

In his 1997 essay “The Cathedral and the Bazaar,” Linux contributor Eric Raymond attempted to explain why open, permissionless infrastructure tends to win in the long run. The accepted wisdom, going back to Fred Brooks’s The Mythical Man Month, was that software had to be built by small, tightly-run teams under a single architect because communication overhead grows quadratically. However, Raymond saw thousands of contributors, most of whom had never met, work on different parts of the Linux kernel simultaneously, and outship billion-dollar companies. If traditional software was carefully crafted like “cathedrals,” the “bazaar” was Raymond’s framing for the chaotic, public, distributed development model that Linus Torvalds had stumbled onto when he made the kernel source code freely available and accepted patches from anyone willing to send them. The guiding philosophy was, in Raymond’s words, “release early and often, delegate everything you can, be open to the point of promiscuity,” and it produced an operating system that, by the early 2000s, ran most of the web.

Raymond’s explanation was that the bazaar avoids the quadratic communication overhead problem because contributors do not coordinate with each other directly. They coordinate with the codebase, through patches and releases, and the maintainership integrates their work into the medium that everyone else then coordinates against. As he put it, “the principle behind Brooks’s Law is not repealed, but given a large developer population and cheap communications, its effects can be swamped by competing nonlinearities that are not otherwise visible.”

Another mechanism Raymond identified was that the bazaar collapsed the distinction between user and developer. In the cathedral, the user is a customer who reports bugs to a service desk. In the bazaar, the user is a co-developer who reports bugs by fixing them, or by describing them in enough technical detail that someone else can. In an open-source community, Raymond explains, “every problem will be transparent to someone.” The crowd works together to outperform any centralized competitor:

“The Linux world behaves in many respects like a free market or an ecology, a collection of selfish agents attempting to maximize utility which in the process produces a self-correcting spontaneous order more elaborate and efficient than any amount of centralized planning could have achieved.”

You can see this happen on Ethereum. Fabian Vogelsteller wrote ERC-20, the standard every stablecoin now uses, because he was building a wallet and there was no clean way to support tokens when each one had a different interface. The ERC-721 standard for NFTs came from the people who made CryptoKitties. Uniswap, now the largest exchange of its kind in the world, started as a blog post by Vitalik Buterin and was built by Hayden Adams, a mechanical engineer with no finance background. None of them needed permission to improve the network. As Sun Microsystems co-founder Bill Joy put it, “No matter who you are, most of the smartest people work for someone else,” and in a permissionless system, innovation can come from anywhere.

What distinguishes the bazaar from the cathedral is that the bazaar’s integration layer is thin, public, and credibility-based rather than authority-based. A coordinator like Linus Torvalds or Vitalik Buterin leads because the contributors choose to follow, and the contributors choose to follow because the coordinator’s decisions can be inspected, criticized and forked if necessary. The Internet has thin centralized integration in the form of the IETF and the IANA. Wikipedia has its editorial processes. Each project that has produced sustained advantage from permissionless innovation combines genuinely open contribution with structured integration that prevents the chaos critics fear. And the integration layer has to operate by credibility rather than coercion or it stops working.

The bazaar also requires a substrate that no one can capture. If Torvalds had tried to take the kernel private, the contributors would have forked the project and continued elsewhere. Raymond developed this idea in “Homesteading the Noosphere,” arguing that open source has developed property rights similar to the Lockean theory of land tenure: a developer establishes ownership of a project by being the first to homestead it (writing the initial code), maintains ownership by continued contribution, and can transfer it through legitimate succession. The credibility of the open license was the formal mechanism, and the noosphere norms were the social mechanism. Take either away and the contributors will go work somewhere else where their contributions cannot be appropriated.

In the Ethereum community, Vitalik Buterin has formalized this requirement under the name “credible neutrality.” A coordination mechanism is credibly neutral when its rules are transparent, when the rules apply equally to all participants, when the rules are hard to change, and when participation is open to anyone willing to follow the rules. These four properties are extracted from observation of systems that have been able to attract contribution at scale. The Internet, Linux, and Wikipedia all have versions of the four properties. Proprietary networks, walled gardens, and corporate blockchains do not.

On a long enough time horizon, credibly neutral systems usually win. The open web displaced proprietary networks; Linux displaced proprietary Unix; Wikipedia displaced Encarta and Britannica. Each time, the proprietary alternatives had real advantages – focused products, more capital, customer support teams, professional marketing and business development teams – and each time, those advantages eroded as the open ecosystem matured and the network effects inverted. Once the open alternative crosses a threshold of accumulated contribution, tooling, and credibility with respect to not changing the rules, it’s almost impossible for a closed system to compete with.

The same pattern is now playing out across every layer of financial infrastructure. SWIFT, Visa and Mastercard, and the consortium chains being pitched to institutions today are different products with different histories, but they’re structurally the same bet: centrally-controlled infrastructure with a latent landlord. For forty years SWIFT was neutral plumbing owned by its member banks until the US pressured it into cutting off Iranian banks in 2012 and several Russian banks in 2022. Despite its corporate governance and Belgian incorporation, SWIFT turned out to answer to the United States, and the rest of the world noticed. China accelerated CIPS, Russia built out SPFS, India expanded UPI, and Brazil’s Pix became the backbone for BRICS Pay. Visa and Mastercard, also born as bank cooperatives, became toll roads charging merchants 1.5-3.5% per transaction. The consortium chains being marketed now (e.g. Canton, Tempo, Arc) carry the identical flaw: a landlord whose interests can diverge from the people building on top of it.

“The original vision of consortium blockchains – the idea that you have 5 banks or major companies that come together and create their own chain – has been mostly a failure,” Vitalik Buterin explains. “It ends up inheriting most of the disadvantages of centralization and most of the disadvantages of decentralization at the same time.” The problem, as he describes it, is that the first few banks feel like equal founders, but bank number twenty is just joining something its competitors already control. You take on all the engineering cost of a distributed system and get none of the benefits of openness, composability, and credible neutrality that made blockchains worthwhile in the first place.

The wreckage backs him up. Between 2017 and 2019, several major bank consortia set out to rebuild trade finance on blockchains. We.trade, backed by a dozen banks including HSBC and Deutsche Bank, went insolvent in 2022. Marco Polo signed more than thirty banks and collapsed into liquidation a year later. Contour shut down months after that. The Australian Securities Exchange spent six years and roughly A$250 million on a permissioned ledger built by Digital Asset (the company now behind Canton) before scrapping it in 2022. Meanwhile Ethereum, which no one controls, has never gone down in its 10+ year history and has only grown.

This is why developers choose Ethereum. By Electric Capital’s count, more than one million developers have contributed to the Ethereum ecosystem over its lifetime, with about 232,000 active in the past year alone. No other chain comes close. Some of that is the ordinary flywheel: the tools, the standards, and the jobs are on Ethereum, so that’s where people learn to build, which attracts more tools and jobs. But developers and institutions are also choosing Ethereum specifically for its superior decentralization and credible neutrality. Last year, for example, Robinhood chose to build a layer 2 on Ethereum rather than their own layer 1; the company’s head of crypto Johann Kerbrat explains the rationale:

“You see a lot of companies right now building their own L1. We were excited about this idea of controlling everything you want to build, but creating the security of a real, proper, decentralized chain is extremely difficult, and you basically get that for free with Ethereum. When you look at some of these new L1s that are being created, it’s not really decentralized and it’s not really secure. At the end of the day it’s basically a fancy database that’s a bit slower to use than an actual database, so we didn’t really see the value in that.”

Erik Voorhees, the founder of Venice AI (the privacy-first AI inference platform with 3+ million users and tens of millions of dollars in ARR), articulated a similar rationale a few days ago. "It wasn't even a question for us," Erik replies when asked why he built Venice on Coinbase's Ethereum L2 Base, "The Ethereum ecosystem is the far more authentic, resilient, and robust ecosystem of all smart contract platforms."

The single-most important blockchain property is sovereignty. What made Bitcoin revolutionary was that it was the world’s first sovereign computer platform. Before Bitcoin, all computer platforms belonged to a person, a company, or a government, and they had to obey the will of their owners and the rules of the jurisdiction where they resided. But a sovereign only obeys its own rules, and no single entity could impose rules on Bitcoin. Kings and queens used to be sovereign, then nation states, and now, for the first time, a computer platform can be sovereign. This is why decentralization is so celebrated in crypto; it is the means by which sovereignty is achieved. A platform with ten validators obeys the rules of those ten. But a platform like Ethereum with hundreds of thousands of independent validators across every major jurisdiction, with multiple independent client implementations, and a foundation that has explicitly disclaimed governance authority, has crossed a threshold where no party can credibly claim ownership. Sovereignty is the property that lets the global financial system build on top of Ethereum without any participant having to fear that another participant, a government, or a foundation can change the rules to their disadvantage.

Much of Ethereum’s lead in sovereignty and credible neutrality comes from path dependence that no other blockchain can replicate. Ethereum launched with proof-of-work in 2015 and ran on it for seven years before transitioning to proof-of-stake in 2022. During that period, ownership of the network was distributed through an open 2014 crowdsale and GPU mining that was deliberately kept accessible to consumer hardware. The result was widespread token distribution with no single entity controlling a meaningful fraction of the network (an essential factor in the sovereignty of a proof-of-stake network). Modern consortium chain launches are venture-funded with concentrated insider allocations, which gives a few actors outsized control over the chain’s consensus. Competitors can copy the architecture but they cannot copy the history.

Since then, Ethereum’s lead has only compounded. The platform’s sovereignty and credible neutrality draws developers. Developers draw more developers because the libraries, tools, and hiring pools that already exist on Ethereum make it easier to build there than anywhere else. Applications draw liquidity and tokenized assets, which in turn draw institutions. Each layer reinforces the others, and a competitor trying to enter has to build all of them at once while Ethereum keeps compounding.

The most sophisticated participants in the space have already chosen Ethereum. Coinbase and Robinhood picked Ethereum for their layer 2s. BlackRock and JP Morgan launched their tokenized money market funds, BUIDL and MONY, on Ethereum. The major DeFi protocols, including Aave, Maker/Sky, Maple, and Uniswap, are predominantly on Ethereum. The largest stablecoin issuers settle on Ethereum. According to Token Terminal’s Ethereum Q1 2026 Report, Ethereum holds 79% of active DeFi loans across the top five chains, 62% of stablecoins, 73% of tokenized funds, and 84% of tokenized commodities.

The applications are permissionless too, which further compounds Ethereum’s advantage. Uniswap’s permissionless listing process, for example, has allowed thousands of long-tail assets to find price discovery and liquidity that no centralized exchange would have provided. Aave’s lending markets are open and composable, which has allowed an entire ecosystem of specialized vaults and risk managers to emerge on top of its liquidity, extending Aave’s reach far beyond what the core team could build alone. Closed systems require gatekeepers to anticipate every use case in advance, but open systems do not.

The strongest objection to this idea “that permissionless wins” isn’t technical; it’s that finance might be the one place where a corporation-owned network is a feature rather than a bug. When a payment fails or an asset ends up somewhere it shouldn’t, a regulator wants someone to answer for it. “Nobody is in charge” sounds less like a virtue and more like a liability when the lawyers show up. But this objection conflates two things that live at different layers. Accountability lives at the application layer rather than the settlement layer. For example, token standards like ERC-3643 embed KYC, identity verification, and jurisdictional transfer restrictions directly into a token’s smart contract so that the issuer can whitelist wallets, restrict transfers, and freeze or claw back assets. Privacy works the same way; zero-knowledge cryptography lets an institution keep transaction details confidential while still settling on the public chain. On a consortium chain, the only parties who can see your data are you and your closest competitors.

In the early days, the Internet was considered far too insecure for real commerce. Then HTTPS made the open web secure enough that essentially all of commerce moved onto it and the question stopped being asked. The skeptics weren’t wrong about the early state of things. They were just wrong about whether an open network could close the gap.

The banks and fintechs building their own chains right now had the same idea AOL and Microsoft had in the early days of the Internet: build the open thing, but inside your own walled garden so you can charge rent. But that never works because the walls that give you control are the same walls that keep innovation out.

The better model is Netscape. Netscape didn’t try to own the web; it built the browser that brought the world onto it. Riding the open web’s explosive growth made it, for a time, one of the most important companies of the era. Ethereum’s credible neutrality is almost impossible to replicate, and it’s already positioned to become the settlement layer for global finance. The winning strategy is to build on top of permissionless infrastructure, not compete with it.

Disclosure: This analysis is published by Etherealize, an organization focused on institutional Ethereum adoption. The author and Etherealize may hold positions in ETH and other digital assets discussed. This is not investment advice.