Ethereum co-founder Vitalik Buterin published a new post, "The cryptographic world computer," on Saturday, laying out what he sees as the network's next decade. His network's next upgrade, he suggested, will be the last one anyone from 2015—the year Ethereum launched—would recognize.

Hegota, the fork scheduled for next year, is "likely to be Ethereum's last 'normal' fork," Buterin wrote. What follows involves recursive STARKs—zero-knowledge proofs able to verify other zero-knowledge proofs—automated formal verification, heavily optimized consensus, and making all of it quantum-safe.

The cryptographic world computer: My attempt to express in somewhat concise terms the true meaning of basically everything planned to happen to Ethereum starting from the fork after Hegota. It's really not just a blockchain anymore. It's a hybrid…

— vitalik.eth (@VitalikButerin) September 27, 2026

The full post is available on vitalik.eth.limo.

A hybrid, not just a blockchain

A post-"Lean Ethereum" network is called a blockchain "to a large extent for historical reasons," Buterin wrote. In reality, he argued, it is a "hybrid construction" merging Satoshi Nakamoto's ideas with cryptographic machinery drawn from 50 years of academic work—machinery that "did not even exist" in 2009. That architecture, he said, will "enable much more powerful properties" than a blockchain alone.

Verification, consensus, and block construction

Three core properties are being replaced outright. Verification moves from downloading and re-executing every block to sampling data with PeerDAS and checking a SNARK—a succinct zero-knowledge proof that stands in for re-executing the block. Consensus shifts to a "much better optimized" proof-of-stake model. Block construction changes from one miner deciding what goes in, to multiple parties building it together.

Final—the point at which a transaction becomes irreversible—would arrive in 8 to 32 seconds, against roughly 200 seconds for 12 confirmations today. Slots would run 4 to 8 seconds, and running a node would demand far less hardware.

Censorship resistance would come from FOCIL—short for fork-choice enforced inclusion lists—a scheme in which a committee of validators each publishes a list of pending transactions that the next block must contain, enforced by the fork-choice rule so that blocks omitting them are ignored. Buterin described the result as "guaranteed real-time transaction inclusion."

Privacy arrives in pieces

Privacy improvements will land incrementally. General-purpose computation remains expensive and only partly private, while special-purpose applications get "very strong privacy" from zero-knowledge proofs and private account abstraction. Onion routing and mixnets, which conceal who is communicating with whom, would cover the network layer—work the Ethereum Foundation resourced with a dedicated privacy cluster launched last October.

Developers would feel the shift in gas costs. The same computation will cost more "if you shove it all into one inscrutable serially-executed transaction," Buterin wrote, and less if it is split into dependencies that can be parallelized or stripped out before the transaction reaches a block.

Decentralization as an advantage

Decentralization is ceasing to be purely a cost, Buterin argued. It lets data be stored in parallel, computation to happen in parallel inside the mempool, and metadata to be hidden in ways a centralized system cannot manage.

That was the mid-2010s dream, he said, and it failed then for one reason: there was no way to verify that split-up work had been done correctly. Committees sampled at random were "complicated to set up and expensive," added latency, and offered "no recourse if the committee fails." Modern cryptography solves the problem, with overhead "decreasing month by month."

Further out sits indistinguishability obfuscation—a way to run programs while hiding how they work—which Buterin describes as the holy grail because it would remove the tradeoff between privacy and generality altogether. He is explicit that nothing in the post depends on it, and that "all of the conclusions in this post will apply long before any of that becomes available."

The harder near-term problem is relatively mundane. Making zero-knowledge proofs fast and safe enough is "encapsulated complexity" already being optimized with AI tools, he wrote. Managing and parallelizing access to very large amounts of state remains "more difficult, and systemically complex."

A companion to the Strawmap

The post reads as a plain-language companion to the Strawmap, the Ethereum Foundation's draft schedule of seven forks through 2029 that Buterin called "a very important document" when Justin Drake published it in February. Where that document lays out dependencies fork by fork, Buterin's post points to the ultimate destination: slot times falling from 12 seconds to two, finality dropping from roughly 16 minutes to single digits, and the whole network replaced component by component in what he then called a "ship of Theseus" rebuild. The next milestone is Hegota itself: once that fork ships next year, the transformation Buterin describes is slated to begin with the fork immediately after it.