I was auditing a DeFi protocol last month when a developer from a prominent staking pool mentioned something that stuck with me: "We don't even ask the Foundation anymore. We coordinate with the other clients." That offhand comment, made over a coffee in Nairobi, captures a transformation that few in the broader market have paused to examine. Ethereum, the network that promised to be a world computer, is undergoing a quiet governance restructuring — one that moves decision-making away from the Ethereum Foundation and toward a diffuse network of client teams, staking pools, infrastructure providers, and large protocol DAOs. This is not a fork or a formal vote. It is a gradual, almost silent power shift, and it carries profound implications for the network's ethos, security, and long-term viability.
Context: The Foundation's Historical Role When I first started writing about Ethereum in 2017, the Ethereum Foundation was the undisputed center of gravity. Vitalik Buterin and a small core of developers guided the protocol's evolution through Ethereum Improvement Proposals (EIPs), funded research, and made pivotal decisions — from the switch to Proof-of-Stake to the parameters of EIP-1559. The Foundation's treasury, then holding over a million ETH, was a war chest for steering the ship. But that model had an inherent fragility: a single point of governance, no matter how benevolent, contradicted the decentralization narrative. The DAO hack, the debate over ProgPoW, and the EIP-1559 conflicts all exposed tensions Under that centralized umbrella.
Today, that umbrella is fraying. The Foundation still holds significant assets — roughly 300,000 ETH as of late 2023 — and funds important research. But its influence on day-to-day protocol decisions is waning. The core developers who implement changes now operate with more independence. Client teams like Geth, Nethermind, Besu, and Erigon each have their own roadmaps and priorities. Staking pools like Lido and Rocket Pool control a meaningful fraction of the validator set. Even infrastructure providers like Infura and Alchemy have become de facto gatekeepers, able to blacklist or prioritize transactions. The result is a multi-node governance landscape where power is distributed — but not necessarily equally or transparently.
Core: The Technical and Ethical Dimensions of Multi-Node Governance Let us examine the architecture of this new governance. In theory, a multi-node structure means that no single entity can unilaterally decide the network's future. In practice, the distribution is lopsided. Geth, for example, runs on over 80% of execution-layer nodes — a dangerous concentration that creates a single point of failure. Should a critical bug appear in Geth, the entire network could stall. Similarly, Lido controls roughly one-third of all staked ETH, giving it outsized influence in any informal governance discussion. Infrastructure dependencies are even more concentrated: Infura and Alchemy together handle the vast majority of RPC traffic, meaning they could, if pressured by regulators, censor transactions or delay upgrades.
From my experience auditing smart contracts in Nairobi, I have learned that "decentralization" is often a story we tell ourselves rather than a technical reality. The multi-node shift does not automatically solve centralization; it replaces a formal hierarchy with an informal one. The real question is whether the new power centers — the client teams, the large staking pools, the infrastructure providers — will act with the same ethical responsibility that the Foundation once held. During the EIP-1559 debate, the Foundation demonstrated a willingness to compromise and listen to community feedback. Can we expect the same from a corporate entity like ConsenSys (which backs Infura) or a liquid staking protocol whose primary incentive is market share?
Tracing the moral code behind every token, we must ask: Who holds the pen when upgrade decisions are made? Currently, the answer is a small group of core developers who meet weekly in the All Core Developers Execution (ACDE) calls. These calls are open to the public, but the real influence often belongs to those who maintain the most widely used clients. The Foundation still convenes these calls, but its authority is increasingly symbolic. Proposals like EIP-4844 (Proto-Danksharding) moved forward largely because client teams, not Foundation executives, championed them. This is both a strength and a vulnerability. It is a strength because it creates redundancy and prevents one person from becoming a dictator. It is a vulnerability because influence can shift to those who control the network's plumbing — without any explicit governance mechanism or accountability.
I recall my work on the ERC-20 standardization in 2017, where I saw how seemingly neutral technical choices could embed systemic bias. The same applies to governance. If the de facto power rests with client developers who are not directly elected or accountable to ETH holders, then we have traded explicit centralization for covert centralization. The Foundation, for all its flaws, had a clear mission and transparent budget reports. The new power centers — the large staking pools, the infrastructure giants — operate with corporate incentives that may not align with the network's long-term health. Building libraries where others build empires means understanding that governance is not just about who votes, but who shapes the options available to vote on.
Contrarian: The Illusion of Decentralization Here is the contrarian angle that most analysts overlook: this silent power shift may actually reduce Ethereum's resilience rather than enhance it. The multi-node narrative sounds like progress, but it masks the emergence of a new oligarchy. Consider the following: If Lido, Geth, and Infura formed an informal alliance — even unintentionally — they could effectively control the network's upgrade path. They could slow-walk competing proposals, prioritize features that benefit their own business models (like cheaper staking or more centralized RPC), and stifle innovation from smaller teams. The Ethereum community is not naive to this risk; there are ongoing efforts to enforce client diversity via software like diversity-fee, and the Foundation has funded alternative execution clients. But the market's inertia is strong. Most users and developers will default to the most popular client or the most liquid staking pool.
Moreover, the loss of the Foundation as a unifying moral authority could lead to more frequent and bitter governance disputes. Without a neutral arbiter, disagreements over contentious EIPs may escalate, leading to network splits or prolonged upgrade delays. We saw glimmers of this during the initial debate over EIP-1559, where miners threatened a fork. Now imagine a scenario where Lido and Rocket Pool disagree on a fee structure change, or where Geth and Nethermind cannot agree on EVM upgrades. The network could stall, and validator coordination could fracture. In such a world, the promise of "code is law" becomes harder to uphold because code is written by people, and people with concentrated power can write laws that favor themselves.
Listening to the silence between the blocks, I sometimes fear that we are lulled by the absence of overt conflict. The quietness of the shift masks the fact that governance is becoming less transparent, not more. When the Foundation made decisions, they published blog posts, held community calls, and defended their choices publicly. Now, decisions are made in the comments of GitHub pull requests by a handful of client maintainers. That is not necessarily bad — technical meritocracy has merits — but it lacks the procedural legitimacy that a broad, diverse user base deserves.
Takeaway: Preserving the Human Story in Digital Ledgers The silent power shift in Ethereum governance is neither a crisis nor a salvation. It is an evolution that demands our attention. If we are to build a truly decentralized financial and social foundation, we must ensure that governance remains accountable, transparent, and aligned with the broader community's values. That means not just celebrating multi-node governance as a buzzword, but actively designing checks and balances — client diversity requirements, staking pool caps, transparent infrastructure policies — to prevent the emergence of a new aristocracy. The Ethereum Foundation may need to reinvent itself as a coordinator rather than a commander, but it should not disappear entirely. Its ethical compass has been a vital asset.
Preserving the human story in digital ledgers means being honest about power, even when it is uncomfortable. The multi-node reality is an opportunity to build a more robust democracy — if we choose to do the work. But if we mistake silence for health, we may wake up to find that we have traded a visible king for an invisible committee. And that, in the end, is no closer to the soul of decentralization.