The most discussed topic in the Ethereum community recently has undoubtedly been Vitalik Buterin's public reflections on the scaling roadmap. Vitalik's attitude can be described as "sharp," stating bluntly that with the improvement of Ethereum's mainnet (L1) scaling capabilities, the roadmap established five years ago, which considered L2 as the primary scaling method, has become ineffective. This statement was initially interpreted negatively by the market as a "pessimistic" or even "negative" view of L2. However, a careful analysis of Vitalik's core viewpoints, combined with Ethereum's mainnet scaling progress, the decentralization assessment framework, and recent technical discussions surrounding Native/Based Rollups, reveals that Vitalik is not completely dismissing the value of L2, but rather aiming for a "correction." Ethereum is not abandoning L2, but rather redefining the division of labor—L1 returning to its role as the safest settlement layer, while L2 pursues differentiation and specialization, thus allowing the strategic focus to return to the mainnet itself. Objectively speaking, in the previous cycle, L2 was indeed once considered Ethereum's lifeline. In the initial Rollup-Centric roadmap, the division of labor was very clear: L1 was responsible for security and data availability, while L2 was responsible for extreme scalability and low gas consumption. In an era when gas costs tens of dollars, this was almost the only feasible answer. However, the reality has been far more complex than expected. The latest statistics from L2BEAT show that there are now over a hundred L2 systems in a broad sense, but this increase in quantity does not equate to structural maturity; the vast majority are progressing slowly in decentralization. Here, it's necessary to add some basic information: back in 2022, Vitalik criticized the training wheels architecture of most Rollups in his blog, stating that it relied on centralized operations and manual intervention to ensure security. Frequent L2Beat users should be very familiar with this; its official website displays a key related metric—Stage. This is an evaluation framework that divides Rollups into three decentralized stages: "Stage 0," which relies entirely on centralized control; "Stage 1," which relies to a limited extent; and "Stage 2," which is completely decentralized. This reflects the degree to which Rollups depend on human intervention in auxiliary wheels. In his recent reflections, Vitalik pointed out that some L2 platforms may remain in "Stage 1" indefinitely due to regulatory or commercial needs, relying on a security council to control upgradeability. This means that such L2 platforms are essentially still "secondary L1" platforms with cross-chain bridging attributes, rather than the "branded sharding" originally envisioned.Or to put it bluntly, if the power to prioritize, upgrade, and make final decisions is concentrated in the hands of a few entities, it not only contradicts Ethereum's original intention of decentralization, but L2 itself becomes nothing more than a parasite leeching off the Ethereum mainnet. At the same time, the proliferation of L2 instances has brought another structural problem that has been deeply felt over the past few years: liquidity fragmentation. This has led to the gradual fragmentation of traffic originally concentrated on Ethereum, forming isolated islands of value. As the number of public chains and L2 instances increases, the degree of liquidity fragmentation will further intensify, which is not the original intention of scaling. From this perspective, it's understandable why Vitalik emphasized that the next step for L2 is not more chains, but deeper integration. Ultimately, this is a timely correction—strengthening L1's position as the most trusted global settlement layer through institutionalized scaling and protocol-inherent security mechanisms. In this context, scaling is no longer the sole objective; security, neutrality, and predictability have once again become Ethereum's core assets. The future of L2 lies not in quantity, but in deeper integration with the mainnet and more specialized innovation in niche scenarios. This could include providing unique additional features such as privacy-focused virtual machines, extreme scalability, or dedicated environments designed for non-financial applications like AI agents. This aligns with the views of Hsiao-Wei Wang, co-executive director of the Ethereum Foundation, at Consensus 2026, who argued that L1 should serve as the most secure settlement layer, handling the most critical activities, while L2 should pursue differentiation and specialization, handling activities that prioritize the ultimate user experience. It is precisely in this wave of reflection on the L2 narrative that the Based Rollup concept is poised for its moment of glory in 2026. If the keyword of the past five years has been "Rollup-Centric," the current discussion is shifting to a more concrete question: Can Rollups "grow within Ethereum" rather than "be attached to outside of Ethereum"? Therefore, the "Native Rollup" currently being hotly debated in the Ethereum community can be understood to some extent as an extension of the concept of Based Rollup—if the native rollup is the ultimate ideal, then the based rollup is currently the most practical path to that ideal. As we all know, the biggest difference between based rollup and traditional L2 such as Arbitrum and Optimism is that it completely abandons the independent, even centralized, sequencer layer, and instead, the ordering is directly performed by Ethereum L1 nodes.In other words, the Ethereum protocol itself integrates Rollup-like verification logic at the L1 level, unifying the extreme performance optimizations and protocol-level security that were previously separate between L2 and the Ethereum mainnet. The most intuitive feeling this design gives users is that Rollups seem to be embedded within Ethereum, inheriting not only the censorship resistance and liveness of L1, but more importantly, solving the most troublesome problem of L2—synchronous composability. Within a Based Rollup block, you can directly access L1 liquidity, achieving atomic cross-layer transactions. However, Based Rollups face a real challenge: if they completely follow the pace of L1 (12 seconds per slot), the user experience will feel cumbersome. After all, under the current Ethereum architecture, even after a transaction is packaged into a block, the system still needs to wait approximately 13 minutes (2 epochs) to achieve finality, which is too slow for financial scenarios. Interestingly, in the same tweet where Vitalik reflected on L2, he recommended a community proposal from January, "Combining preconfirmations with based rollups for synchronous composability." The core of this proposal isn't simply promoting Based Rollups, but rather a hybrid structure: retaining low-latency sequenced blocks, generating based blocks at the end of a slot, submitting the based blocks to L1, and finally combining this with a preconfirmation mechanism to achieve synchronous composability. In Based Rollups, preconfirmation involves a specific role (such as an L1 proposer) promising that the transaction will be included before it is formally submitted to L1. This is precisely what Ethereum's Project #4, Fast L1 Confirmation Rule, explicitly outlines in the Interop roadmap. 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 13 minutes required for full finality. From a mechanistic perspective, the fast confirmation rule doesn't introduce a new consensus process, but rather reuses the attester voting that occurs in every slot of the Ethereum PoS system. When a block has accumulated enough validator votes in an early slot, even before entering the finalization stage, it can be considered "extremely unlikely to be rolled back under a reasonable attack model." In short, this level of confirmation doesn't replace finality, but rather provides a strong confirmation explicitly acknowledged by the protocol before finality. This is particularly crucial for Interop. Cross-chain systems, Intent Solvers, and wallets no longer need to blindly wait for finality; instead, they can securely advance to the next step of logic within 15–30 seconds based on protocol-level confirmation signals.Through this layered confirmation logic, Ethereum has finely differentiated different levels of trust between "security" and "perceived speed," potentially creating an extremely smooth interoperability experience (further reading: "Ethereum's 'Second-Level' Evolution: From Fast Confirmation to Settlement Compression, How Does Interop Eliminate Waiting Time?"). Looking back from the 2026 juncture, Ethereum's main theme is quietly shifting, gradually moving from pursuing ultimate "scalability" to pursuing "unity, layering, and intrinsic security." Last month, several executives of Ethereum's L2 solutions expressed their willingness to explore and embrace the Native Rollup path to improve the consistency and coordination of the entire network. This attitude itself is an important signal: the Ethereum ecosystem is undergoing a painful but necessary de-bubbling process, returning from pursuing "the number of chains" to pursuing "protocol unification." However, as Ethereum's underlying roadmap is recalibrated and advanced, especially as L1 continues to be enhanced and Based Rollup and preconfirmation are gradually implemented, and underlying performance is no longer the only bottleneck, a more realistic problem begins to emerge—the biggest bottleneck is no longer the chains, but wallets and entry barriers. This confirms the insight imToken repeatedly emphasized in 2025: as infrastructure becomes increasingly invisible, the true limit of scalability will be the entry-level interactive experience. Overall, besides underlying scaling, the future expansion and scalability of the Ethereum ecosystem will not focus solely on TPS or Blob counts, but will revolve around three more structurally significant directions: Account abstraction and barrier-of-sight elimination: Ethereum is promoting native account abstraction (Native AA), and future smart contract wallets will become the default choice, completely replacing obscure mnemonic phrases and EOA addresses. For wallet users like imToken, this means entering the crypto world will be as simple as registering a social media account. Privacy and ZK-EVM: Privacy features will no longer be a peripheral need. With the maturity of ZK-EVM technology, Ethereum will maintain transparency while providing necessary on-chain privacy protection for commercial applications. This will be its core competitive advantage in the public chain competition. On-chain sovereignty of AI agents: In 2026, the initiator of transactions may no longer be a human, but an AI agent. The challenge of the future lies in establishing trustless interaction standards: how to ensure that AI agents are acting on the user's will, rather than being manipulated by a third party? Ethereum's decentralized settlement layer will become the most reliable rule-maker for the AI economy. Returning to the initial question, did Vitalik truly "deny" L2? A more accurate understanding is that he rejected an overblown, fragmented narrative that is detached from the mainnet and operates independently.This is not the end, but a brand new beginning. Moving from the grand vision of "brand sharding" to the meticulous refinement of Based Rollups and pre-confirmation essentially helps strengthen Ethereum L1's absolute position as the foundation of global trust. However, this also means that in this return to technological pragmatism, only those innovations truly rooted in the underlying principles of Ethereum's new phase and sharing the same fate as the mainnet will survive and thrive in the next great age of exploration. [imToken]