Vitalik: Ethereum Solves the Scalability Trilemma

Ethereum’s trilemma: solved (in code, at least)

Vitalik Buterin is saying something pretty wild: after a few big upgrades in 2025, Ethereum may have finally squashed the old “scalability trilemma” — the idea that you can only pick two out of decentralization, security, and scalability. He points to data availability sampling (the PeerDAS stuff) hitting mainnet and production-quality ZK-EVMs as the tech combo that lets Ethereum be decentralized, secure, and high-bandwidth at the same time.

If you like internet metaphors, think of BitTorrent as a massive, chaotic party with tons of bandwidth but no single source of truth, and Bitcoin as a super-strict librarian where everyone copies the same book slowly. The new Ethereum, according to Vitalik, is trying to be both the party and the librarian: lots of throughput but with consensus and verification that actually mean something.

Before you pop the champagne, the ZK layer is described as “alpha but usable” — it gets the job done, but safety checks and wider node support are still rolling out. Data availability tech is already running, which makes the whole idea feel less like a theoretical paper and more like real, running software.

Roadmap, risks, and the walkaway test

Vitalik’s timeline lays out the next few years: expect big gas-limit increases and non-ZK improvements around 2026, the emergence of ZK-EVM node runners starting that year, state and gas repricing work through 2026–2028, and a heavier reliance on ZK validation by 2027–2030. The idea is a gradual shift from “every node does all the work” to “nodes verify compact proofs,” which should let the network carry way more traffic without selling out on decentralization.

There’s also a long-term nerdy dream called distributed block building — basically, make sure no single place ever constructs a whole block. Whether that happens inside the protocol or via distributed marketplaces, the aim is to reduce gatekeepers and make transaction inclusion fairer across geography and infra setups.

But here’s the ideological curveball: Vitalik isn’t just celebrating engineering wins. He’s warning that these tools are only meaningful if the community uses them to build resilient, genuinely decentralized apps — not just to chase the next meme coin or squeeze clicks out of tokenized dollars. To measure success he proposes the “walkaway test”: can an app keep working without its creators around? If not, it hasn’t passed.

So the big question for the next decade isn’t whether the code can scale — it’s whether people will use the scaling to make a true “world computer” that survives abandonment, censorship, and rent-seeking intermediaries, or whether the tech will get co-opted into the same centralized, subscription-first internet we were trying to escape. Spoiler: the tech’s ready for a rebellion. The community’s job is to decide what kind of rebellion it wants.