Vitalik Buterin: The original vision of Ethereum L2 is no longer applicable; a new path needs to be found.

Vitalik Buterin, in an article published on the X platform, stated that there has been increased discussion recently regarding the continued role of L2 in the Ethereum ecosystem. Due to the much slower-than-expected progress of L2 into Phase 2, and the ongoing scaling of L1 itself, with a significant increase in gas limits anticipated in 2026, the original vision of L2 as a "branded shard" for Ethereum is no longer applicable, necessitating a new path. He pointed out that L1 no longer needs L2 as a branded shard, and L2 is unable or unwilling to meet the attributes required for true branded sharding. He suggested that L2 should identify value-added aspects beyond "scaling," such as privacy, application-specific efficiency, extreme scaling capabilities, non-financial application design, ultra-low latency, and built-in oracles. If processing ETH or other Ethereum assets, L2 should at least reach Phase 1 and support maximum interoperability with Ethereum. Furthermore, Vitalik Buterin stated that he has become more convinced of the value of native Rollup pre-compilations in recent months, especially after acquiring the ZK-EVM proofs required to scale L1. This pre-compilation will make EVM verification without a security committee feasible. He believes that the design of this pre-compilation should be explored so that it can verify the EVM portion when L2 contains "EVM and other content". This would facilitate secure, robust, and trustless interoperability with Ethereum and enable synchronous composability.