Vitalik Says Most L2s Aren’t Really Ethereum

Ajouté : 2026-05-11

[00:00:00]Vitalik Buterin just told the layer two ecosystem that most of them aren't actually scaling Ethereum. They're building separate blockchains with multi-sig bridges and calling it something that it is not. Now this matters cuz we're talking about 30 billion dollars worth of rollups that process the majority of all Ethereum transactions. Now of course the responses were immediate and revealing as well. Arbitrum's co-founder fired back that >> [music] >> Arbitrum is not Ethereum. A notable shift from their 2024 messaging and Optimism CEO acknowledged that they're not ready for full trustlessness. And about two weeks later though for separately stated reasons Base [music] announced that it is leaving the Optimism Superchain to build its own stack. So, let's break down what Vitalik actually said, why the major L2 teams are pushing back and also what this means for where the ecosystem goes from here.

[00:01:00]So here is the core of Vitalik's argument and it's more nuanced than the headlines made it sound. He wrote, [music] "If you create a 10,000 TPS EVM where its connection to L1 is mediated by a multi-sig bridge, then you are not scaling Ethereum." [music] Now the original promise of layer twos was that they would be extensions [music] of Ethereum itself. You would get cheaper transactions but your funds would still be secured [music] by Ethereum's consensus. The rollup would post proofs to Ethereum and if anything went wrong, the proof system would [music] catch it. But most L2s haven't achieved that. They have fraud proofs or validity proofs but there is still a group of humans with multi-sig keys who can [music] pause the chain, upgrade contracts and intervene. So, when Vitalik says most L2s aren't [music] scaling Ethereum, he is pointing out a gap between how these chains market themselves and what they actually technically provide. If your security depends on a committee of humans rather than Ethereum's consensus, [music] you're trusting a different set of people. But Vitalik's critique also has another layer. And this one might be even more uncomfortable for L2s. Even if they solve the decentralization problem, Ethereum itself might not need them for basic scaling anymore. Think about why L2s became necessary in the first place.

[00:02:32]Back in 2020 and 2021, Ethereum was genuinely struggling. Gas fees would spike to over hundreds during peak congestion. If you wanted to use Ethereum for anything routine, you either had deep pockets or you needed another option. But Ethereum has been scaling its base layer directly as per its roadmap. Today a simple transfer on mainnet costs a few cents, not far away from many L2s. Now of course this is also during a bear market [music] but if Ethereum L1 becomes cheap enough for most use cases, then what's the point of an L2 that's just Ethereum but [music] cheaper? The cost advantage shrinks and if that L2 is also still relying on a security council rather than trustless proofs, you're paying similar fees for weaker security guarantees.

[00:03:25]>> [music] >> So that is why Vitalik argued that L2s now need to offer something Ethereum physically cannot provide. Something different like privacy built into the execution model, extreme throughput in the tens of thousands of TPS or ultra-low latency with millisecond block times. Now the responses from major L2 teams reveal how they are thinking about their position. Steven Goldfeder from Arbitrum pushed back directly stating [music] that Arbitrum is not Ethereum.

[00:03:57]Now that is a pretty [music] big shift from the messaging that he used in 2024 when he posted things like Arbitrum is Ethereum. [music] Now he's drawing a clearer distinction. Arbitrum is a close ally but [music] it is its own thing.

[00:04:13]His substantive argument was about peak [music] demand highlighting how Arbitrum and Base were each processing over 1,000 [music] transactions per second during recent market volatility while Ethereum L1 was handling around 40. He added that even with scaling upgrades [music] L1 cannot absorb sudden spikes. He also warned that if Ethereum becomes hostile to rollups, institutions might build [music] their own L1 chains instead.

[00:04:43]Karl Floersch from Optimism accepted Vitalik's challenge while being candid [music] about limitations. He acknowledged that full trustlessness isn't production ready and that week-long withdrawal delays are a real UX problem. [music] Jesse Pollak from Base agreed that L2s can't just be Ethereum but cheaper going forward and they need to offer unique value like [music] supporting different applications, account abstraction, privacy and [music] more. And then about two weeks after Vitalik's post, Coinbase announced Base is leaving the OP stack to build [music] its own infrastructure.

[00:05:22]And their stated reasons were about development velocity, not Vitalik's critique. Now the timing is notable but we can't claim that that is why they left. Vitalik's whole critique comes down to this gap between what L2s promise and what they actually deliver.

[00:05:41]So, where do the major chains actually stand right now? L2Beat has long had a staging framework to measure this. Stage two is the finish line, fully code controlled, no human override, users get at least 30 days to exit before any upgrade takes effect. Projects like Cartesi's Honey Pot and ZK Money qualify but combined they have less than 3 million in them. So, for practical purposes there are no major rollups at [music] stage two. Most of the value sits at stage one which means the proofs work but a security council can still override them. The big optimistic rollups are here. Arbitrum with roughly 15 billion, Base with about 10 billion and OP mainnet around 2 billion. Then there's stage zero where the rollup is effectively run by operators and some well-known chains such as Linea and ZK Sync Era are still here. Part of the slow progress is genuinely technical but part of it is also incentives. For some L2s staying at stage one is actually the business model and Vitalik acknowledged this directly. He said some L2s may never want to go beyond stage one because their customers' regulatory requirements demand ultimate control.

[00:07:04]And his take was that this might be the right thing for those customers but it means you're not scaling Ethereum in the sense the rollup-centric roadmap originally meant. And that is fine he argued because Ethereum is now scaling L1 directly anyway. So, [music] where do we go from here? Well, some chains will push towards stage [music] two and genuine trustlessness. Some will stay at stage one with clear acknowledgement that it is a different security model and some will build genuinely differentiated capabilities that justify their existence beyond cost saving. That said, the L2s at the top [music] right now, Arbitrum, Base, Optimism are thriving with billions in value and millions of users and many of those users probably already know about the security council tradeoffs and use them anyway. The ecosystem that they've built, the applications, the liquidity has real value. The Ethereum but cheaper positioning is getting harder to maintain as Ethereum itself gets cheaper but whether that actually reshapes which chains people use remains to be seen.

[00:08:16]These networks have momentum >> [music] >> and momentum still matters. But what happens when Ethereum mainnet itself is genuinely fast and cheap [music] enough for everything? Will L2s lose relevance?

[00:08:30]Well, give us your thoughts in [music] the comments below. Bye.