Vitalik's Speech in Hong Kong: The Future Five Years and Long-Term Vision of the Ethereum Protocol

On the morning of April 20th, Ethereum co-founder Vitalik Buterin delivered a speech at the opening ceremony of the 2026 Hong Kong Web3 Carnival. Vitalik Buterin comprehensively outlined Ethereum’s five-year technical roadmap, clarifying Ethereum’s core value positioning, phased technology upgrade plans, and long-term foundational principles.

In his speech, Vitalik stated that Ethereum has two irreplaceable core values, which also serve as its foundational support as Web3 infrastructure:

• First, as a decentralized public bulletin board, providing a trusted on-chain data publishing and notarization backbone for various applications, ensuring transparency and verifiability of information, immutability of order, and equal rights for anyone to publish information. It is the core technical support for scenarios like privacy protocols, secure electronic voting, and more;

• Second, as a trusted shared computing platform, supporting shared digital objects automatically executed by code, covering all categories such as ERC-20 tokens, NFTs, ENS domains, DAOs, etc., providing decentralized applications (DApps) with autonomous, secure, verifiable, and fair participation capabilities.

Vitalik also emphasized that the most valuable long-term Web3 applications are based on a deep integration of on-chain and off-chain components, rather than simple on-chain replication of traditional applications.

Regarding short-term technical implementation plans for Ethereum over the next one to two years, Vitalik elaborated on four core upgrade directions:

• Continuous on-chain scalability iterations: Ongoing optimization of Gas Limit, promotion of CKEVM deployment, and implementation of series of EIP proposals to enable block parallelization and gas re-pricing, enhancing on-chain data publishing and complex computation capacity while maintaining network security;

• Full deployment of account abstraction: Advancing the implementation of EIP-8141 account abstraction proposal, enabling native support for smart contract wallets, compatible with pay-for-others transactions, quantum-resistant signature algorithms, privacy protocols, etc., greatly expanding Ethereum’s application scope;

• Proactive layout for quantum resistance: Addressing potential risks of quantum computing by optimizing hash-based and lattice-based quantum-resistant signature algorithms, and upgrading EVM vectorization to solve current issues of low efficiency and high on-chain resource consumption of quantum-resistant signatures, preparing Ethereum’s post-quantum security foundation;

• Privacy and storage capacity breakthroughs: Continuing to strengthen on-chain privacy support, focusing on overcoming technical challenges in storage expansion, and addressing Ethereum’s storage capacity limitations.

Below is the full transcript of Vitalik Buterin’s speech:

Good morning everyone! Where is Ethereum headed? I believe we’ve seen many significant changes in the fields of theory and ecosystem. Over the past few years, we’ve also witnessed many changes outside the Ethereum ecosystem, including the possibilities brought by artificial intelligence, the imminent arrival of quantum computing, advances in formal verification, cryptography, and zero-knowledge proofs.

I think one of the many things we’ve been doing is rethinking what truly matters — what is the significance of using this protocol? What features does this protocol have? Why does a random variable need to possess these features? For example, how to integrate it into the Ethereum protocol we previously established, and what will be Ethereum’s use cases in the next five years? I believe it has two main functions:

First, the protocol is like a public billboard, a place where applications can publish messages, and everyone can see the content and order of these messages. These messages can be anything — transactions, hashes, encrypted data, or many other things. In fact, applications have many opportunities to leverage Ethereum as a data publishing platform, interpreting this data through other types of protocols — decrypting data and performing computations on it.

Second, computational capability: Ethereum essentially provides shared digital objects controlled by code, in various forms — assets, ERC-20 tokens, NFTs, ENS domains, DAOs, etc. Their significance is not just theoretical; ENS is a typical example. They can even represent control over organizations, with DAOs being a prime example. We can use them to realize many functions, making these two core features highly valuable. For decentralized applications, they ensure autonomous security, verifiability, and fair participation, bringing all users together.

Sovereignty essentially means that as a user, you can participate, verify, and ensure your own security, all based on your own infrastructure. You don’t need to trust any third party to run the protocol, nor do you need to trust parties outside the protocol, if you choose not to.

Therefore, verifiability and validation capabilities are crucial — they ensure the correct operation of the chain, verify that everything on-chain complies with rules, and also safeguard everyone’s right to publish information, allowing anyone to post on this bulletin board. That’s the core: We should see the protocol as a technical module, and think about all applications that this technical foundation can support. The most interesting applications are inevitably products of on-chain and off-chain integration, including ENS, prediction markets, etc. Prediction markets have on-chain components, such as tradable assets created for each event; and off-chain components, like oracles.

Additionally, sometimes the design of prediction markets or order matching occurs on-chain, involving privacy-related aspects. For example, for decades, cryptographic protocols have been studied to simplify or realize secure electronic voting. Many such protocols rely on public bulletin boards for publishing information. In these scenarios, people publish encrypted votes, ensuring everyone can participate. Any privacy-related matter must include an on-chain component for publishing data and an off-chain component for interpreting it.

Interpreting this data requires off-chain private protocols. That’s why we often talk about Layer-2 (L2). I define meaningful L2 as two types: meaningless and meaningful — meaningless L2 simply copies a protocol and amplifies throughput by 100 times, ultimately becoming more centralized; meaningful L2 requires us to review applications first, clarify what off-chain components are involved, what additional parts are needed beyond FRAML, and then build it.

What does this mean for the protocol? We need to expand data, enhance on-chain data publishing capacity, and improve peer-to-peer networks — this has been reflected in recent hard forks, included in last year’s updates, but we still need further progress. Computing power expansion is equally important because it’s part of Ethereum’s chain, helping different applications integrate and communicate directly without intermediaries.

You can visit Strong Map Organization, which has a roadmap designed for the next five years. The protocol’s short-term core goals are twofold: first, short-term scalability — continuously and actively increasing gas limits; second, launching CKEVM. CKEVM allows Ethereum to support more scenarios and perform more complex computations while still easily verifying on-chain information. Additionally, we are working on early preparations for the post-quantum era — for years, we’ve been paying attention to quantum computing, aware of the challenges it will bring, and have already devised some countermeasures. In the short term, we will improve quantum safety protocols and optimize the entire roadmap.

Our ultimate goal is to make the protocol fully quantum-safe, ensuring all parts are secure and efficient, while improving modular construction processes and strengthening privacy support. Therefore, many EIP proposals for short-term scalability will be implemented in subsequent phases: after protocol hard forks, block access lists can enable parallel processing; gas re-pricing can improve efficiency and security, making higher gas limits safer.

EPBS extends the time required for verifying protocols and blocks, further enhancing security, and also boosting the protocol’s capabilities — supporting state download via EIP-8141, an account abstraction proposal that is simple yet powerful. Essentially, a transaction is a series of calls, where one call might be verification, another execution, enabling native support for smart contract wallets, used for paying others’ transactions, and supporting quantum-resistant signatures and privacy protocols.

This broadens Ethereum’s use cases and supports more functions. Quantum-resistant signatures have existed for 20 years; we understand their principles and how to build them. But the issue is efficiency: a four-round signature consumes 2000 to 3000 bytes, while current signatures only need 64 bytes; on-chain consumption can reach 200k gas, but now only about 3,000 gas. So, we plan to adopt two types of signatures: hash-based signatures and lattice-based signatures. Our approach is to add vectorization capabilities into the EVM, using principles similar to those that enable fast AI computations — in the past, we processed large data sets in parallel; this logic can also be applied to quantum-resistant signatures, striving to make signatures both quantum-resistant and more efficient.

Account-level storage, account balances, and smart contract execution are relatively easier to implement, but storage expansion remains challenging — there’s still much work to do in this area.

That’s our plan for Ethereum’s short-term and long-term development, and the direction we hope Ethereum will evolve.

Ethereum is not competing with high-frequency trading platforms, nor is its goal to be the fastest chain. Its goal is to be a secure, decentralized chain — a chain that stays online and can be trusted. One of our aims is to maximize secure consensus: if the network is secure, it can withstand 49% node failures, and even operate normally when almost all nodes are offline, providing security comparable to Bitcoin. Even if the network encounters issues, we can still maintain 33% security certainty — that’s the first part of our goal.

The second part is formal verification of everything. We’ve already begun actively using AI-generated data proofs to verify that long-term software versions of the protocol have the required properties. We’ve made progress that was impossible two years ago. AI is advancing rapidly, and we’re leveraging this advantage, pursuing simplicity to prepare for the future.

Therefore, a protocol needs to pass an “exit test” — if a protocol is to be deployed practically, it must be dependable, capable of running even without power (no available electrical outlets). This aligns with Bitcoin’s goal and what we must achieve: to safeguard digital assets long-term, we need to build a system that provides continuous security, whose security does not depend on any team’s ongoing existence or work. Lean consensus combines the strengths of both approaches: one is the Bitcoin-style global chain, and the other is BFT (finality), which offers optimal security, quantum resistance, and fast finality.

Thus, finality can be achieved within 1 to 3 slots, with an expected total confirmation time of about 10 to 20 seconds, or even less. zkVM (Zero-Knowledge Virtual Machine) enables you to verify the correctness of the chain without relying on large computers running all operations — everyone should be able to verify the chain before trusting it, even on your phone or IoT devices. Currently, zkVM speeds are fast enough to support real-time virtual machine proofs. Our goal this year is to ensure its security, deploy zkVM on a small portion of the network initially, and gradually increase its application. By 2028, through zkVM deployment, Ethereum will be capable of scaled expansion, handling more transactions while maintaining decentralization.

What is the vision behind all this? Ethereum is a world computer. It is both a global shared layer for making commitments, publishing data, recording actions, and enabling different user operations; and a platform accessible to everyone, where data published or not can be proven. At the same time, it is a global shared layer for ensuring the execution of high-value rules — a protocol that must be highly robust and easily verifiable. I believe that in the future, with the help of AI, ensuring software security will become easier and more straightforward than we imagine.

If you want to ensure software security but neglect safety issues, the number of software vulnerabilities and attacks will be ten times higher than before. Therefore, as a blockchain, Ethereum must prioritize security first, then decentralization; once these are guaranteed, we should do our best to provide this security to users. If you aim to build decentralized applications, you need to ensure their sovereignty, verifiability, and user participation rights. This covers fields like finance, decentralized social, identity verification, and more — including financial and non-financial applications like ENS and prediction markets. The core goal of the protocol is to make application development simple, which is also our default effort.

The roadmap for the next four years is designed around this goal. Thank you!