Is L2 completely dead, or has it taken a turn for the better? To summarize Vitalik's words, Ethereum will no longer pin its hopes on L2 scaling. The future direction is: L1 becomes the trusted base layer and nativeizes ZK-EVM verification capabilities (through native rollup pre-compilation), using mathematical correctness rather than security committees to guarantee state validity. L2 will no longer be responsible for "scaling Ethereum", but will instead serve as a functional layer around Ethereum, providing differentiated capabilities such as privacy, non-EVM, ultra-low latency, social networking, identity, and AI. Currently, without star products, L2 is basically beyond saving; but only when products become dependent on this chain can an L2 be pulled from "no one uses it" back to "having value". For example, Polygon's revenue from its prediction market, Polymarket, also exceeded its base. Vitalik's words foreshadow the complete end of the era of L2 games that were gas-depleted. It's quite a pity. The massive scale of OP and ARB attracted many people to contribute to the blockchain, increase monthly active users, and generate large amounts of data, keeping Ethereum on the chain. At that time, we lived a fulfilling life and were full of hope for the future. Even so, there is no need to be pessimistic. Opportunities are often brewing in bear markets, and this applies to short selling as well. A bear market means: competition decreases significantly, projects return to their true value, and truly promising projects can still secure good funding even in a bear market. Unlike a bull market, where any project can have a story and raise money.