Vitalik Buterin Announced That Blockchains Are Shifting to a New Computing Model

• Vitalik Buterin believes that advances in cryptography will allow blockchains to move away from a model where every node re-executes the same computations.
• SNARKs, STARKs, and other technologies are expected to boost both network scalability and privacy at the same time.
• In the developer’s view, AI will speed up the development and formal verification of more complex blockchain systems.

Ethereum co-founder Vitalik Buterin spoke at the 12th Global Blockchain Summit as part of Shanghai Blockchain International Week. He explained how advances in zero-knowledge proofs, formal verification, and artificial intelligence are reshaping blockchain architecture.

第十二届区块链全球峰会正在进行中！

@ethereum 联合创始人 @VitalikButerin 线上发表《从区块链，到区块链 + 密码学》主旨演讲，探讨区块链与密码学深度融合的技术演进方向，展望隐私保护与安全计算的更多可能！

观看直播https://t.co/7wt8wHuI41 pic.twitter.com/kYP9wRErDQ

— Wanxiang Blockchain (@WXblockchain) September 23, 2026

According to Buterin, over the past 15 years, not only have the industry’s technical capabilities changed, but so has the understanding of what properties decentralized systems should provide. 

Initially, blockchain made it possible to determine who owns assets and who has the right to manage them. The next stage will be programmable privacy, where code also determines who can access specific information.

He noted that early blockchain architecture required many participants to repeatedly perform the same operations. This approach ensured verifiability, but limited performance and privacy.

The development of SNARKs and STARKs is gradually changing this model. Instead of re-executing computations, one participant will be able to perform an operation and generate a cryptographic proof, while everyone else will only verify its correctness.

“The era when everyone has to repeat everything is coming to an end,” Buterin said in his talk.

STARKs Can Change Ethereum’s Architecture

The developer named recursive STARKs as one of the key focus areas. They make it possible to combine many proofs into one, reducing the amount of data and computation that needs to be processed directly on the blockchain.

Previously, we reported that Buterin proposed using this approach as part of EIP-8288. The initiative envisions aggregating cryptographic signatures and STARK proofs at the mempool level, followed by publishing a compact proof of their validity.

Buterin expects the technology to reduce the cost of private and post-quantum transactions, and to move some complex cryptographic operations off Ethereum’s main execution path.

This approach is tied to a broader change in the network’s transaction structure. In early September, the Ethereum co-founder spoke about EIP-8141, which splits a transaction into separate actions and dependencies. This architecture should make it possible to process some operations in parallel instead of executing them sequentially.

In his new talk, Buterin also drew attention to FOCIL — a mechanism for distributed block building. Its goal is to reduce reliance on a single participant when including transactions in a block, and to improve the network’s resistance to censorship.

In the long run, the developer believes the blockchain should be viewed not only as an L1. It will represent a broader compute pipeline that includes local operations on the user’s device, mempools with different privacy characteristics, proof aggregation, L2s, and the network’s base layer itself.

Buterin expects this architecture to significantly increase scalability without the need to fully abandon decentralization guarantees.

In the summer of 2026, the developer already spoke about Ethereum’s transition to a notional “third era,” in which the network’s main priorities will include scaling, privacy, and resilience to quantum computing.

AI Will Accelerate the Development and Verification of Blockchain Systems

Buterin assigned a separate role in this transformation to artificial intelligence. According to him, modern cryptographic systems are becoming so complex that manual development and code audits are gradually turning into a limiting factor.

AI can be used to write and analyze software, create specifications, run tests, and perform formal verification. The latter makes it possible to mathematically prove that an implementation meets the stated requirements.

In Buterin’s view, combining several independent specifications with automated verification will make it possible to build significantly more complex systems without a proportional increase in risk.

He also linked the shortcomings of today’s L2s to limited capabilities for verifying complex code. Projects are forced to use additional security mechanisms, including multiple proof systems and security committees.

Earlier, Buterin said that combining AI and formal verification could significantly reshape Ethereum’s security. In particular, algorithms can help find vulnerabilities, verify protocol implementations, and match code against formal specifications.

At the same time, AI’s use is not limited to development. Buterin believes agents will be able to interact independently with complex infrastructure spanning L1, L2, and intermediary services, choosing the right tools to carry out a specific operation.

The intersection of the two technologies remains one of the areas the Ethereum co-founder is paying increased attention to. In February 2026, he presented an updated vision for the development of AI and Ethereum, emphasizing decentralization, privacy, and verifiability.

In closing, Buterin urged developers to rethink their approach to building applications. In his view, many existing products were designed during a period when advanced cryptography, programmable privacy, and automated formal verification were not yet considered practically viable.

Now these constraints are gradually fading, opening the door to a new class of applications built on Ethereum and other decentralized networks.

Сообщение Vitalik Buterin Announced That Blockchains Are Shifting to a New Computing Model появились сначала на INCRYPTED.