Ethereum was not created to make finance efficient or apps useful. It is designed to free people.
That line from the Trustless Manifesto was criticized when it was published, and Vitalik Buterin repeated it on January 5.
The argument: Ethereum’s mission is fundamentally different from the efficiency game in which DeFi protocols compete. The goal is not a return of 4.5% versus 5.3%, not reducing latency from 473 milliseconds to 368, and not reducing login from three clicks to one.
Ethereum’s game is resilience: avoiding total losses when infrastructure collapses, governments turn hostile, or developers disappear. Resilience means keeping the 2,000 millisecond latency at 2,000 milliseconds even if Cloudflare fails, sponsors go bankrupt, or users are removed from the platform.
Resilience means remaining a first-rate participant, regardless of geography or politics.
This is important because Ethereum is anchoring nearly $74 billion in smart contract value only in layer-1, and more than 65% of tokenized real-world assets.
Yet the system designed as the world’s computer sits on a surprisingly fragile pile of centralized bottlenecks.
The consensus protocol continued to finalize blocks, but the RPC provider’s outdated client caused exchanges to crash. The blockchain continued to run, but the CDN went dark, causing half of the ecosystem to go offline.
Catastrophe avoidance over yield optimization
A recent report quantifies the stakes: Infrastructure failures cause volatility shocks 5.7 times larger than regulatory announcements for major crypto assets. The tail risk of total loss of access, permanent blocking of funds and network shutdown is more important than the increasing returns.
A protocol that offers a 5.3% return is worthless if a configuration error can destroy the infrastructure.
Vitalik Buterin’s framing captures this. Resilience is not about speed when everything works, but about whether your application will still work at all if infrastructure providers disappear or if hosting platforms deplatform users.
The 2000 millisecond latency that Ethereum delivers may be slower than Web2, but it continues to deliver even if Web2 systems shut down completely.
Still, Ethereum’s resilience promise faces practical tests.
In November 2020, Infura, the default RPC provider for MetaMask and most DeFi apps, was running an outdated Geth client that deviated from the canonical chain.
Exchanges halted Ethereum withdrawals, Explorers exposed conflicting states, and MakerDAO and Uniswap broke for users.
While the bug itself has been fixed and progress is being made on alternative RPC implementations, centralization remains the norm. It’s just less Infura-only and more”small cartel.”
The protocol worked, but the attachment points failed.
In November 2025, a Cloudflare configuration error disabled approximately 20% of web traffic, including Arbiscan, DefiLlama, and multiple exchange and DeFi front ends. Ethereum continued to process blocks. Users could not access it.
During the 2024 inscription craze, Arbitrum’s single sequencer stuck for 78 minutes. No transactions processed, no batches posted to Ethereum.
Arbitrum, Optimism, Base, and zkSync all currently rely on single, centralized sequencers. The decentralized base layer functioned correctly, but the centralized infrastructure prevented users from taking advantage of it.
| Low | Current dependency | Fragility metric | Resilient alternative |
|---|---|---|---|
| Access / RPC | Infura, Alchemy, QuickNode; MetaMask is Infura by default | ~90% of Web3 app traffic; November 2020 An Infura glitch halted ETH withdrawals, broke MetaMask, MakerDAO, Uniswap | Multiple RPC providers, local light clients, standard stateless clients; RPC diversity as a user-centric feature |
| Relay / Builder | MEV Boost relay (Ultra Sound, Titan, bloXroute) that mediates >90% of the blocks | Four relays control >85% of proposals; Titan, Beaverbuild and Rsync produce >80% of the building blocks | More relays by different entities; transmit neutrality; anchored PBS where relay faults cannot block the block space |
| L2 Sequence | Some sequencers (Arbitrum Foundation, Optimism Foundation, Coinbase for Base) | Arbitrum: 78 minutes downtime; Base captures 70.9% of L2 profits, Arbitrum 14.9%, Optimism 5.4% | Decentralized sequencer sets or L1 fallback; forced recording when sequencer censors; track % L2 TVL under one control |
| DNS/CDN | Cloudflare for DNS, TLS, dApp caching | Cloudflare ~20% of the global internet; An outage in November 2025 disabled Arbiscan, DefiLlama, exchange/DeFi front-ends | IPFS/Arweave with ENS fallbacks; multi CDN; wallets that call contracts without a web front end |
| Basic protocol | Ethereum Consensus (Lighthouse 52.65%, Prysm 17.66%); execution (Geth ~41%, Nethermind 38%) | September 2025 Reth bug stopped 5.4% of nodes; diversity prevented a broader impact | No customer >33% share; deportation at home; minimize correlated failure; simple light/stateless client authentication |
The core protocol demonstrates true resilience, with multiple clients, hundreds of thousands of validators, and proof-of-stake that spreads risk across codebases.
When Reth encountered a bug in September 2025, it blocked 5.4% of nodes, but network continuity was maintained as Geth, Nethermind, and Besu continued. Customer diversity worked.
The problem centers above: RPC access, relays, sequencers, and web front ends introduce dependencies that disable user access even if the base layer is functioning.
This is where Ethereum’s resilience breaks: not in cryptography or consensus, but in the scaffolding that connects users to the protocol.
Centralized sequencers as economic bottlenecks
Layer-2 sequencers concentrate both control and gain. Base consistently achieved more than 50% of all total profits in 2025, followed by Arbitrum.
Arbitrum’s sequencer is managed by the Arbitrum Foundation, Optimism’s by the Optimism Foundation, Base’s is centralized by Coinbase, and zkSync’s.
As a result, more than 80% of the fees captured by Ethereum Layer-2 in 2025 flowed to blockchains with centralized sequencers.
The technical path exists: shared sequencer networks like Espresso, or based on rollups that return sequencing to Ethereum validators. Astria tried similar designs but stopped in 2025.
The gap is not technical, but economic. Centralized sequencers deliver better UX and generate significant revenue. Resilience requires you to accept that a sequencer that produces slightly slower confirmations, but is impossible to turn off by a single operator, will beat millisecond improvements with single-point control.
RPC and CDN dependencies
MetaMask is Infura by default. Reports note that most Web3 applications use Infura, Alchemy or QuickNode.
The November 2020 Infura incident demonstrated the consequence: protocol-level resilience became irrelevant when the access layer failed.
Cloudflare’s November 2025 outage revealed how much “decentralized finance” depends on a company’s CDN. Ethereum processed blocks normally, but users were unable to reach front ends, explorers, or dashboards.
Resilient alternatives include wallets that use multiple RPCs by default, local lightweight clients, distributed storage on IPFS or Arweave, ENS addressing, and multi-CDN deployments.
However, these come with costs such as increased complexity, higher bandwidth requirements and more complex management.
Most projects opt for convenience, and therefore the trade-off between efficiency is important. Ethereum’s base layer provides survivable features, while the ecosystem largely wraps these in dependencies that reintroduce any vulnerability.
The actual interaction
Ethereum’s value proposition, as Buterin puts it, is not faster, cheaper or more convenient. It works when everything else breaks.
That requires infrastructure choices that prioritize survival over optimization: multi-client deployments if one is technically superior, diverse RPC providers if one offers better latency, decentralized sequencers if centralized operators deliver faster acknowledgments, and distributed front ends if centralized hosting is simpler.
The sector has not embraced this trade-off. Optimize rollups for UX and accept the risk of a single sequencer. By default, applications use convenient RPCs and accept concentration risks. Front ends are deployed on commercial CDNs and tolerate single-vendor failures.
The choice: build for the case where Cloudflare, Infura, and Coinbase all continue to work, or build for when they don’t.
Ethereum’s base layer makes the second choice possible. The surrounding ecosystem overwhelmingly takes first place.
The protocol offers 2,000 milliseconds of latency that persists through infrastructure failures, deplatforming and geopolitical disruption.
Whether someone builds systems that actually leverage this feature rather than wrapping it in dependencies that reintroduce every vulnerability Ethereum was designed to eliminate will determine whether resilience becomes real or remains theoretical.
Block space is plentiful. Decentralized, permissionless, resilient block space is not.