The core assumptions of blockchain are being rewritten.

Written by: Vitalik Buterin

Translated by: Saoirse, Foresight News

Editor’s Note: This article is an expanded written version of Vitalik's keynote speech at the Shanghai Blockchain International Week. Based on the speech, the author has added a large amount of comparative analysis and technical tables, systematically explaining how Ethereum is evolving from a traditional ledger to a "cryptographic world computer," analyzing underlying transformations such as PoS, ZK proofs, PeerDAS, etc., looking ahead to the technical roadmap after the Hegota fork, and exploring a new balance paradigm of decentralization, privacy, and scalability. Click to view Vitalik's speech content at Shanghai Blockchain International Week.

We refer to Ethereum as "a blockchain," as if it essentially belongs to the same category of technology as the Bitcoin created by Satoshi Nakamoto in 2009. In many ways, the two indeed resemble each other, even the "lightweight Ethereum" planned according to the Strawmap in the future retains the core characteristics of blockchain. However, at the same time, this technology has undergone immense evolution over the past fifteen years and will continue to iterate over the next three years. By that time, it would be quite reasonable to refer to the form of Ethereum that is evolving as a system of an entirely different nature.

Today's Ethereum has general computing, a proof-of-stake mechanism, on-chain applications powered by zero-knowledge proofs, and layer two networks that achieve scalability and privacy protection. In the future, Ethereum's computing power can flexibly adjust between extreme scalability and full generality; it will have various multi-party participation models for block construction, deep-optimized proof-of-stake, and zero-knowledge proofs will be integrated into the underlying protocol to play a core role.

This article will outline some of the most crucial fundamental differences between the blockchain of the 2010s and the blockchain of 2030 from a technical perspective and the system characteristics accessible to users.

First, we will study the original Bitcoin white paper section by section, comparing it with Ethereum's differences, using yellow to represent Ethereum in 2015, green for 2025, and blue for 2030.

This image is adapted from the transaction principle diagram of Satoshi Nakamoto's Bitcoin paper, showing the classic chain-like signature transfer model and noting that the new solution can aggregate signatures off-chain, submitting only a single record on-chain, often utilizing zero-knowledge proofs to replace traditional signatures.

This image compares Bitcoin's PoW mechanism with Ethereum's PoS after 2022: where blocks were previously produced by a single participant iteratively searching for a qualifying hash through nonce, now it will replace that with validators' signatures, and in the future, it will utilize FOCIL for multi-role decentralized block construction, while transaction-related signatures, proofs, and other components will be split and aggregated in the memory pool.

This image excerpts the six-step process of network operation from the Bitcoin white paper, indicating that its native design lacks capabilities such as memory pool aggregation, distributed block construction, and sender anonymization, while also illustrating the improvements of Ethereum's PoS and PeerDAS: separating block construction and fork selection, where nodes only need to download a small portion of the block, relying on parallel proofs to reduce consensus delay.

This image compares the Merkle tree pruning scheme proposed in the Bitcoin white paper with Ethereum’s new storage strategy: Bitcoin can delete spent transactions to free up hard disk space, while in the future, Ethereum will further reduce storage through the separation of state and history, SNARK proofs, and will also introduce distributed state storage and various storage media optimization.

This image interprets the privacy concept from the Bitcoin white paper: relying solely on public key anonymity can only obscure identity, while transaction amounts remain public, which, under modern data analysis, is insufficient; however, ZK-SNARK, FOCIL, and EIP-8288 can construct more robust programmable privacy, while also addressing the problems of querying reading privacy and network broadcast privacy.

Almost every chapter of the white paper will exhibit significant changes. To streamline the content, we have organized it into the following table:

Almost all core attributes encompassed by the concept of blockchain have either fundamentally changed or are about to undergo underlying transformations:

• Verification Mechanism: Download and re-execute → PeerDAS sampling and SNARK verification
• Consensus Mechanism: Proof of work → Proof of stake → Deeply optimized proof of stake
• Block Construction Authority: Single miner generates blocks → Multiple parties jointly construct blocks

In the voice of AI, the only objective and credible conclusion (well, it is the core viewpoint): modern cryptographic networks like the streamlined Ethereum are called "blockchain" largely for historical reasons. In reality, it is a hybrid architecture that integrates two major systems:

1. Satoshi's core concepts
2. A whole new class of powerful cryptographic tools born from fifty years of academic research, many of which either did not exist or were not mature in 2009

How much cryptography is hidden in the cryptographic system? Comparison of the years 2009, 2020, and 2030

Cryptography is not the only key discipline. Formal verification, database theory, improvements in P2P network theory, information theory, economics, and other fields are also equally important. But all these technologies can accommodate this foundational logic: everyone attempts to generate the next block containing a valid proof of work, once one person successfully generates it, they broadcast it, and all others download the block and re-execute, repeating the cycle. The transformations at the cryptographic level are entirely different.

So what does all this mean for users? The most important conclusion is that the dimensions of trade-offs users need to consider are undergoing a radical transformation:

When developing applications, computational structure becomes crucial. In a simple blockchain, one byte equals one byte, and one unit of Gas equals one unit of Gas. However, in the future architecture: if you pack all computations into a single transaction that is difficult to disassemble and sequentially executed, the cost for the same amount of computation will be much higher; but if you split the computations into well-packaged, parallel-supportive or pre-emptive independent tasks, preferably completing preprocessing before the transactions are input into the final block, the cost will significantly decrease. This will change a developer's incentive orientation; over time, the architecture of all Ethereum applications will follow suit: in the long run, we may develop a new programming paradigm — only the core information describing non-exchange state changes and execution order will be put on-chain, while all other data will be aggregated before being included in the block.

Reasonably organizing computation can allow blockchains to focus more on their core tasks.

Perhaps the most significant shift is that the decentralization attribute of the network is no longer merely a performance burden taken on to exchange for safety and robustness; in certain limited scenarios, decentralization itself can even become a performance advantage.

Decentralized networks can store vast amounts of data in parallel; massive computations can be executed in parallel, with many calculations completed directly in the transaction memory pool. In certain contexts, decentralization can also enhance privacy because only decentralized networks can effectively hide metadata (such as the sources of data and requests).

As early as the mid-2010s, this was Ethereum's early vision: decentralization is not solely for enhancing robustness but also to improve scalability. Centralized systems can enhance performance by splitting tasks among multiple participants, and blockchains can do the same. At that time, this idea could not be realized, primarily due to the verification mechanism. After task splitting, you must verify that each sub-task has been correctly executed. Early solutions attempted to solve the problem with random sampling committees but faced the same bottleneck: first, committee deployments are complex, costly, and significantly increase delays; second, once a committee fails, there are no remedies. Now, relying on modern cryptography, this problem has been solved, and the additional overhead from this solution is decreasing monthly.

Another noteworthy direction where decentralization is expected to improve performance is latency. Ethereum's inherent latency can never match that of centralized servers, but infrastructures built on Ethereum can achieve that.

Overall, building a powerful decentralized middle layer (which is not blockchains) between users and the main chain can greatly enhance Ethereum's capabilities while not undermining the core characteristics of the underlying blockchain.

Looking further into the future, Ethereum may also undergo a new wave of transformation — program obfuscation (iO) may rise. The ultimate goal in this field: mature and usable obfuscation techniques can eliminate the trade-off between privacy and generality. You can achieve fully general computing involving an unlimited number of (asynchronous) participants in a completely securely encrypted form. Even a weakened version of obfuscation technology has many practical scenarios, such as encrypted transaction memory pools. However, all conclusions presented in this article hold true before this technology is implemented.

This is a cryptographic world computer: Ethereum is no longer just a ledger where developers can freely write computing tasks and data to be executed; the new architecture integrates blockchain, cryptographic privacy, cryptographic verification, along with powerful decentralized off-chain components into one.

Fully implementing this design still presents many challenges. Making zero-knowledge proofs efficient and secure is not easy, but it belongs to encapsulated complex problems, which are currently being massively optimized with AI tools. The more daunting and systematically complex issues likely involve managing and parallel accessing vast states. Many solution ideas have already emerged, but they need to be continuously refined, especially as we further understand what applications will be run in the future.

Looking at the Strawmap roadmap, it can be seen that the upcoming Hegota hard fork planned for next year is likely Ethereum's last "regular hard fork," its functionalities and technologies still familiar to developers from 2015. All upgrades after Hegota will include recursive STARKs, automated formal validation, highly optimized consensus algorithms, and an entire system resistant to quantum changes. The realization of PeerDAS marks Ethereum's initiation of transformation: evolving from a simple blockchain to a significantly more capable system. After the Hegota upgrade, this transformation will become the mainline of Ethereum's development. The ultimate goal: to achieve a highly secure computational system that is cheaper, more scalable, and offers better privacy protection compared to the previous generation of technology. This is the cryptographic world computer.