Recently, the most discussed topic in the Ethereum community has undoubtedly been Vitalik Buterin's public reflection on the scaling roadmap.
It can be said that Vitalik's attitude was quite "sharp,"直言随着以太坊主网(L1)自身扩容能力的提升,那个在五年前制定的、将 L2 视为主要扩容手段的路线图,已经失效了。
These remarks were once interpreted negatively by the market as "pessimism" or even "denial" towards L2. However, if one carefully examines Vitalik's core points and considers Ethereum's series of mainnet scaling progress, the decentralization progress evaluation framework, and recent technical discussions around Native/Based Rollup, it becomes clear that Vitalik is not completely negating the value of L2, but rather engaging in a kind of "course correction":
Ethereum is not abandoning L2, but is redefining the division of labor—L1 returns to its role as the most secure settlement layer, while L2 pursues differentiation and specialization, thereby refocusing the strategic emphasis on the mainnet itself.
一、Has L2 Fulfilled Its Historical Mission?
Objectively speaking, in the previous cycle, L2 was indeed once seen as Ethereum's lifeline.
In the initial Rollup-Centric roadmap, the division of labor was also very clear: L1 was responsible for security and data availability, L2 was responsible for extreme scaling and low Gas fees. In an era where Gas fees often reached tens of dollars, this was almost the only feasible answer.
But reality developed in a more complex way than expected.
The latest statistics from L2BEAT show that there are now over a hundred L2s in the broad sense, but quantitative膨胀 does not equal structural maturity, with the vast majority making slow progress in decentralization.
Here, some basic knowledge needs to be added. As early as 2022, Vitalik criticized the "Training Wheels" architecture of most Rollups in a blog post,直言其依赖中心化运维、人工干预保障安全. Frequent users of L2Beat should be very familiar with this, as its homepage displays a key related indicator—Stage:
This is an evaluation framework that categorizes Rollups into three stages of decentralization: "Stage 0" which fully depends on centralized control, "Stage 1" with limited dependence, and the fully decentralized "Stage 2". This also reflects the degree of reliance on the artificial intervention of training wheels.
In his recent reflection, Vitalik pointed out that some L2s, possibly due to regulatory or commercial needs, might forever remain at "Stage 1", relying on security councils to control upgradability. This means that such L2s are essentially still a "secondary L1" with cross-chain bridging properties, rather than the "branded shards" initially envisioned.
Or, to put it more bluntly, if the sequencing power, upgrade rights, and final arbitration are concentrated in the hands of a few entities, it not only runs counter to Ethereum's decentralized初衷, but the L2 itself is nothing more than a parasite leaching off the Ethereum mainnet.
At the same time, the proliferation of L2s has brought another structural problem that everyone has felt deeply over the past few years: liquidity fragmentation.
This has caused the流量 originally gathered on Ethereum to be gradually divided, forming一个个被割裂的价值孤岛. As the number of public chains and L2s increases, the degree of liquidity fragmentation will intensify further, which was not the original intention of scaling.
From this perspective, one can understand why Vitalik emphasizes that the next step for L2 is not more chains, but deeper integration. Ultimately, this is actually a timely course correction—through institutionalized scaling and protocol-inherent security mechanisms, strengthening L1's positioning as the world's most trusted settlement layer.
Against this background, scaling is no longer the sole goal; security, neutrality, and predictability have重新成为以太坊的核心资产. The future of L2 lies not in quantity, but in deeper integration with the mainnet and more specialized innovation in细分场景.
For example, providing unique additional functions, such as privacy-specific virtual machines, extreme scaling, or specialized environments designed for non-financial applications like AI agents.
The views of Ethereum Foundation Executive Director Hsiao-Wei Wang at Consensus 2026 coincide with this: L1 should serve as the most secure settlement layer, carrying the most critical activities; while L2 should pursue differentiation and specialization, carrying activities that追求极致用户体验.
二、Native Rollup: Based Rollup + The Future of Preconfirmations?
It is in this wave of reflection on the L2 narrative that the concept of Based Rollup is expected to have its moment of glory in 2026.
Because if the keyword of the past five years was "Rollup-Centric," then the core of the current discussion is shifting to a more specific question: Can Rollups be "grown inside Ethereum" rather than "attached outside Ethereum"?
Therefore, the "Native Rollup" currently hotly debated in the Ethereum community can, to some extent, be understood as an extension of the Based Rollup concept—if Native Rollup is the ultimate ideal, then Based Rollup is currently the most practical path towards that ideal.
As is well known, the biggest difference between Based Rollup and traditional L2s like Arbitrum and Optimism is that it completely抛弃了独立的、甚至是中心化的序列器(Sequencer)层, and instead is directly sequenced by Ethereum L1 nodes. In other words, the Ethereum protocol itself integrates Rollup-like verification logic at the L1 level, thus unifying the extreme performance optimization and protocol-level security that originally belonged to L2 and the Ethereum mainnet separately.
The most intuitive feeling this design gives users is that the Rollup seems to be embedded within Ethereum, inheriting not only L1's censorship resistance and liveness but, more importantly, solving the most headache-inducing problem for L2s—synchronous composability. In a Based Rollup block, you can directly call L1's liquidity, achieving atomic cross-layer transactions.
However, Based Rollup faces a practical challenge: if it completely follows L1's rhythm (one Slot every 12 seconds), the user experience would be cumbersome. After all, under the current Ethereum architecture, even after a transaction is packed into a block, the system still needs to wait about 13 minutes (2 Epochs) to reach finality. This is too slow for financial scenarios.
Interestingly, in the same tweet where Vitalik reflected on L2, he recommended a community proposal from January titled "Combining preconfirmations with based rollups for synchronous composability". The core of this proposal is not simply pushing Based Rollup, but rather proposing a hybrid structure:
Retain low latency sequenced blocks, generate a based block at the end of the slot, submit the based block to L1, and finally combine it with a preconfirmation mechanism to achieve synchronous composability.
In Based Rollup, a preconfirmation is a commitment made by a specific角色 (like an L1 proposer) that a transaction will be included before it is formally submitted to L1. This is also what Project #4: Fast L1 Confirmation Rule in Ethereum's Interop roadmap explicitly proposes to do.
Its core goal is very direct: to allow applications and cross-chain systems to obtain a "strong and verifiable" L1 confirmation signal within 15–30 seconds, without having to wait the full 13 minutes required for finality.
Mechanically, the fast confirmation rule does not introduce a new consensus process, but rather reutilizes the attester voting that occurs in every slot of Ethereum's PoS system. When a block has accumulated enough, and sufficiently dispersed, validator votes in early slots, even if it has not entered the finalization phase, it can be considered "extremely unlikely to be reverted under a reasonable attack model".
Simply put, this confirmation level does not replace Finality, but provides a strong confirmation explicitly acknowledged by the protocol *before* Finality. For Interop, this is particularly critical: cross-chain systems, Intent Solvers, and wallets no longer need to blindly wait for finality; instead, they can safely advance the next logic within 15–30 seconds based on a protocol-level confirmation signal.
Through this layered confirmation logic, Ethereum finely segments different levels of trust between "security" and "perceived speed,"有望构建出极致丝滑的互操作体验 (extended reading: "Ethereum's 'Second-Level' Evolution: From Fast Confirmations to Settlement Compression, How Interop Eliminates Waiting Time?").
三、What is Ethereum's Future?
Looking back from the vantage point of 2026, Ethereum's main theme is quietly shifting, gradually moving from pursuing extreme "scaling" to pursuing "unification, layering, and inherent security."
Last month, executives from several Ethereum L2 solutions have successively expressed their willingness to explore and embrace the Native Rollup path to enhance the consistency and synergy of the entire network. This attitude itself is an important signal: the Ethereum ecosystem is undergoing a painful but necessary deflation of泡沫, shifting from pursuing "the number of chains" back to pursuing "the unification of the protocol."
However, as Ethereum's underlying roadmap is recalibrated and advanced, especially as L1 continues to strengthen and Based Rollup with preconfirmations are gradually implemented, and底层性能不再是唯一瓶颈时, a more practical problem begins to emerge—the biggest bottleneck is no longer the chain, but the wallet and the entry barrier.
This confirms the insight that imToken repeatedly emphasized in 2025: when infrastructure becomes趋于隐形, what truly determines the limit of scale will be the entry-level interaction experience.
Overall, besides底层扩容, the future breakthrough and规模化发展 of the Ethereum ecosystem will not focus solely on TPS or Blob count, but will revolve around three more structurally significant directions:
- Account Abstraction and the Erosion of Entry Barriers: Ethereum is promoting native account abstraction (Native AA). Future smart contract wallets will become the default choice, completely replacing obscure seed phrases and EOA addresses. For users of wallets like imToken, this means the barrier to entering the crypto world will become as simple as registering a social media account (extended reading: "From EOA to Account Abstraction: Will Web3's Next Leap Happen in the 'Account System'?");
- Privacy and ZK-EVM: Privacy features are no longer边缘需求. As ZK-EVM technology matures, Ethereum will provide necessary on-chain privacy protection for commercial applications while maintaining transparency. This will be its core competitive advantage in the public chain competition (extended reading: "The 'Dawn Moment' of the ZK Route: Is Ethereum's Endgame Roadmap Accelerating Comprehensively?");
- On-Chain Sovereignty of AI Agents: In 2026, the initiator of transactions might not be humans, but AI agents. The future challenge lies in establishing trustless interaction standards: how to ensure that AI agents are executing the user's will, rather than being manipulated by third parties? Ethereum's decentralized settlement layer will become the most reliable rule arbiter for the AI economy (extended reading: "The New Ticket in the Age of AI Agents: What is Ethereum Betting on by Promoting ERC-8004?");
Returning to the initial question, did Vitalik really "deny" L2?
A more accurate understanding is that he denied an overly膨胀、脱离主网、各自为政的碎片化叙事. This is not an endpoint, but a brand new starting point. From the grand illusion of "branded shards" back to the meticulous crafting of Based Rollup and preconfirmations, this本质上反而有助于强化以太坊 L1 作为全球信任底座的绝对地位.
However, this also means that in this return to technological pragmatism, only those innovations that are truly rooted in the underlying principles of Ethereum's new phase and share the same breath and fate as the mainnet will survive and thrive in the next great navigation era.
