Ethereum's Vitalik Buterin says blockchains are moving away from every node repeating all computations. New cryptographic tools like zero-knowledge proofs and FHE could let networks check results without redoing all the work.
Vitalik Buterin, co-founder of Ethereum, wants to end a core blockchain habit: making every node repeat every computation. At Wanxiang Blockchain Week on September 23, 2026, Buterin said the old model of "everyone re-executes everything" is fading. He sees a future where advanced cryptography lets networks check results without all the extra work. This could change how blockchains scale, protect privacy, and run complex apps.
From redundant execution to proof
Blockchains like Bitcoin and Ethereum have always used brute-force consensus. Every node runs every transaction again, checking signatures, balances, and state changes to stop cheating. This makes the system easy to verify, but it's costly. As blockchains take on more complex tasks, every participant has to process the full load. That limits how much the network can grow and forces all data and logic out in the open.
Vitalik Buterin's 2026 speech marked a pivotal shift: Ethereum and other blockchains are moving from full re-execution by every node to proof-based verification, leveraging SNARKs and STARKs for efficiency and privacy.
Buterin's latest comments mark a turning point. Instead of just upgrading hardware or tweaking code, he asks a basic question: does every node really need to re-execute every computation to check if it's right? If not, blockchains could let specialized actors do the heavy work, while the rest of the network just checks cryptographic proofs. This is already happening with zero-knowledge (ZK) tech. One participant does the computation and creates a proof. Others only need to verify it. That cuts out the need for everyone to repeat the same work. This is how SNARKs and STARKs work, and they're now key for ZK-rollups and new scaling solutions as detailed by CoinDesk.
Zero knowledge and the rise of proof systems
Zero-knowledge proofs (ZKPs) are driving this change. With SNARKs and STARKs, one party does the work and creates a short proof that the result is correct. Other nodes don't have to repeat the work-they just check the proof. This is already powering ZK-Rollups, where thousands of transactions happen off-chain and only a proof goes to the main blockchain. Multiple independent reports say Ethereum plans to use more STARKs over the next two years. ZK infrastructure is seen as the backbone for future upgrades in scaling and privacy.
Buterin said these proof systems are moving from research to real use. Ethereum's roadmap includes proposals like EIP-8288, which would combine proofs and signatures to cut on-chain verification costs. The direction is clear: blockchains are becoming layered networks. Computation, proof generation, and verification are split between different roles, not forced on everyone. Industry commentary now describes blockchains as cryptographic networks for secure computation, with a focus on privacy, formal verification, and proof aggregation at the block level.
Ethereum's transition to proof-based verification is underpinned by its proof-of-stake consensus, but the majority of mainnet activity still relies on full node execution. As SNARKs and STARKs mature, Ethereum's roadmap anticipates broader adoption of these technologies for both scalability and privacy.
Programmable cryptography and expanding control
The impact goes beyond scaling. Programmable cryptography, as Buterin puts it, expands blockchain security from "who can send what asset" to "who can see what information, where and how computation happens, and how correctness is proven." Fully homomorphic encryption (FHE) lets data stay encrypted even while it's being processed. Zero-knowledge proofs let users show they're eligible or compliant without giving up private details. This is a big change from early blockchains, which focused almost only on asset ownership and transfers. Recent reports link Buterin not just to ZK, but to this wider idea-programmable cryptography as the next step after basic blockchain transparency.
As blockchains take on identity, enterprise data, and privacy-heavy uses, programmable cryptography becomes key. It gives fine control over who can access data and how computation runs, all enforced by code and cryptography. The result is a blockchain world that can be both transparent and private, depending on what users and apps need.
Formal verification and AI in blockchain engineering
Buterin also pointed to formal verification and AI as ways to make these cryptographic systems practical. As smart contracts and proof systems get more complex, manual audits aren't enough. Formal verification uses math to prove code does what it's supposed to do. AI is starting to help with code generation, testing, and optimization. This could make advanced cryptographic tools easier for more developers to use. Industry reports say AI-driven formal verification is becoming a practical way to secure cryptographic and smart contract systems.
For more on how privacy and computation are changing in DeFi, see EgonCoin's recent coverage of encrypted trading protocols.
Looking forward, Buterin sees blockchains as cryptographic networks, not just ledgers. User devices, privacy layers, specialized proof generators, and base chains will each have their own jobs. The main chain will move from re-executing everything to just checking what's been computed elsewhere. This could cut costs, boost privacy, and allow richer on-chain apps-if the engineering hurdles can be cleared.
Ethereum's move to proof-based verification isn't just a technical tweak. It's a shift in how trust, privacy, and computation are balanced on public blockchains. As programmable cryptography, ZKPs, FHE, formal verification, and AI come together, blockchains are moving away from redundancy and toward efficient, flexible verification. For developers, this means rethinking how execution and validation are split up. For users, it could mean lower fees, stronger privacy, and new kinds of apps that weren't possible before.
On September 23, 2026, Ethereum's network processed transactions using its current proof-of-stake consensus. All validators still had to check state transitions themselves. SNARKs and STARKs are already used in some scaling solutions, but most of Ethereum's mainnet activity still depends on full node execution. As of that date, no major blockchain had fully switched to a proof-only model, but research and pilot projects are picking up speed across the industry.
Zero-knowledge proofs and fully homomorphic encryption each solve different parts of blockchain privacy and scaling. ZKPs let one party prove a computation is correct without showing the data. FHE lets computation happen on encrypted data. Both cut down on the need for public transparency, but they come with big engineering and performance trade-offs. As these tools get better, developers will have to balance privacy, efficiency, and verifiability when building new blockchain apps and protocols.
