At the Hong Kong Web3 Festival, Vitalik Buterin once again brought up a familiar question: what exactly should Ethereum become?

Over the past few years, the crypto industry has grown used to discussing public chain competition through performance metrics. A chain with higher TPS, faster confirmations, and lower gas fees is often more easily understood as “next-generation infrastructure,” and so on.

But by 2026, both Ethereum and new public chains are facing more than internal crypto topics such as DeFi, NFTs, L2 scaling, and on-chain finance. AI Coding is rapidly absorbing attention and resources, while formal verification and zero-knowledge proofs are also making clear progress.

All of this means the world public chains face is changing again. Do we still need public infrastructure that anyone can verify, exit, use for self-custody, and rely on without being controlled by a single point of failure?

And how, exactly, is Ethereum thinking about and preparing for that future?

1. What Is Ethereum Actually For: A Bulletin Board and Computation

“Ethereum is not trying to compete with high-frequency trading platforms. Ethereum is not trying to become the fastest chain. Ethereum is meant to be a secure chain, a decentralized chain, a chain that stays online, and a chain you can always rely on.”

In this speech, Vitalik used two foundational concepts to explain Ethereum’s value again: first, Ethereum is like a “public bulletin board”; second, Ethereum provides “computation.”

These two simple concepts are not only aligned with the new positioning of Ethereum that we have been discussing recently, but also capture the fundamental reason Ethereum is different from ordinary internet services.

This “public bulletin board” is not an abstract metaphor. It means applications can publish messages on Ethereum, and everyone can see both the content and the order of those messages. These messages can be transactions, hashes, encrypted data, or other information that needs to be publicly committed, ordered, and verified. Further reading: From “Global Computer” to “Bulletin Board”: What Do Ethereum and Vitalik Actually Want to Build?

This is also the most fundamental difference between Ethereum and ordinary servers. Servers can be faster, cheaper, and more efficient, but they usually require users to trust that the operator will not tamper with records, deny service, or shut down the system at a critical moment. Ethereum is designed to provide that capability without relying on this kind of trust.

Computation, meanwhile, means Ethereum allows people to create shared digital objects controlled by code. These objects can be ERC-20 tokens, NFTs, ENS names, DAOs, on-chain organizations, financial protocols, or other more complex applications.

So by 2026, if we still understand Ethereum only through “performance,” or compare it with new public chains only by TPS, gas costs, and confirmation speed, we may easily miss the real problems Ethereum is trying to solve next.

Blockchain scaling over the next few years is no longer just about “making chains faster.” It is about preserving verifiability, decentralization, and user self-sovereignty across more complex use cases. In other words, scaling is not about turning Ethereum into another centralized high-performance system. It is about enabling more applications to run without sacrificing the trust assumptions of the base layer.

This is also a key reason why Vitalik has been rethinking L2s since the beginning of this year.

Put simply, he believes the market has long treated L2s mainly as Ethereum’s scaling tools: when the mainnet becomes expensive or congested, more transactions move to L2s. But today, the first-stage mission of L2s has largely been completed. They should no longer remain merely at the level of “offloading transactions.” Instead, they should become the frontier where Ethereum expands into more application scenarios.

This point is especially important for today’s Ethereum ecosystem.

Over the past few years, the market has often simplified L2s as “a cheaper Ethereum.” But in Vitalik’s framework, L2s are not simple substitutes for Ethereum. They are functional extensions built around Ethereum’s public base layer. L1 carries the most critical roles of commitment, settlement, data publication, and verification, while L2s and off-chain systems provide higher-frequency, more flexible, and more privacy-friendly execution capabilities for specific use cases.

This is why Vitalik does not want “being the fastest” to become Ethereum’s primary goal.

Speed is certainly important. But if speed comes at the cost of ordinary users being unable to run nodes, verify state, or protect themselves when something goes wrong, then the chain will gradually become a less efficient centralized service.

For Ethereum, speed is a user experience issue. Security and decentralization are the reason it exists.

2. In the AI Era, Ethereum’s Value May Be Amplified Again

One of the most noteworthy parts of this speech is that Vitalik did not treat AI merely as an external hot topic. Instead, he placed it inside the technical context of Ethereum’s future roadmap.

For example, the Ethereum community has already begun experimenting with using AI to generate proofs for code, in order to show that Ethereum software versions have the properties they are supposed to have. Two years ago, this would still have been difficult. But the rapid development of AI is making software security verification easier.

Behind this is a very practical issue: as blockchains carry more assets, identities, organizations, and rules, the cost of code vulnerabilities will become higher and higher. If AI can help developers discover vulnerabilities, generate proofs, and assist with formal verification, then it is not only an efficiency tool at the application layer. It may also become part of protocol security engineering. Further reading: When Hackers Scale Up with AI: The Next Level of Web3’s Security Arms Race

But AI’s deeper impact on Ethereum is not only on the developer side. It also affects users, especially by changing how ordinary users interact with digital systems.

Human-computer interaction has gone through several major shifts over the past few decades. In the earliest stage, users interacted with computers through command lines, and only a small number of technically skilled people could truly use complex systems. Later, graphical interfaces and mobile apps became mainstream, allowing ordinary users to complete actions through buttons, pages, and menus.

Now, AI is shifting the paradigm toward natural language interaction. Users no longer need to understand every step. They only need to state their goal, and the system may automatically break down the path, call the necessary tools, and complete the execution.

In Web3, this shift could have an even greater impact.

Today, if a user wants to complete a cross-chain DeFi operation, they often need to choose a network, confirm gas fees, approve contracts, execute a swap, bridge assets, and then deposit into a protocol. Every step requires a signature, and every step may go wrong.

In the future, if AI Agents become an important entry point for wallets and on-chain applications, the user may only need to say: “Swap part of my ETH into stablecoins and deposit it into a yield protocol under a low-risk strategy.” The remaining path planning, protocol selection, transaction simulation, and execution may all be handled by an intelligent agent.

This sounds like it could significantly lower the barrier to using Web3. But it also raises an important problem. When users no longer click through every step themselves, and when AI Agents interpret intent, call contracts, and initiate transactions on users’ behalf, how can users confirm that the agent has not exceeded its permissions? How can they know the selected path is not malicious? How can they retain the ability to revoke, verify, and protect themselves without sacrificing user experience?

This is exactly where Ethereum’s value is amplified again.

AI can make interactions more natural, but natural language itself does not create trust. If a smarter interface is still backed by an unverifiable black-box system, users are merely shifting from “trusting the platform” to “trusting the model.” Ethereum, by contrast, is closer to trust infrastructure for the AI era.

Taking this one step further, this also makes the role of wallets more important. Future wallets may no longer be just “signing tools” or “asset lists.” They may gradually become the permission management layer between users and AI Agents, on-chain applications, identity systems, and payment networks.

Users will need wallets to define boundaries: which actions can be executed automatically, which actions require additional confirmation, which assets cannot be used, and which approvals need to be reviewed and revoked regularly.

3. CROPS: From Foundation Mandate to Community Contract

Interestingly, shortly before Vitalik re-explained Ethereum from the perspective of the protocol roadmap, the Ethereum Foundation also released the EF Mandate, which can be seen as a formal confirmation of this direction at the level of values.

The Mandate states that the ultimate reason Ethereum exists is to protect users’ self-sovereignty: users should be able to independently control their assets, identity, actions, and choices without relying on any centralized intermediary.

Around this idea, the EF Mandate introduces the acronym “CROPS”: Censorship Resistance, Open Source, Privacy, and Security. In the Foundation’s wording, Ethereum must first preserve these four properties. Without them, Ethereum loses the reason why it is worth using, worth building on, and worth protecting.

To be fair, censorship resistance, open source, privacy, and security are not new terms in the Web3 context. They have been discussed repeatedly since the early days of the crypto industry. But emphasizing them again today carries a clearly different meaning.

After all, when the early crypto industry discussed these values, it was mostly in opposition to centralized platforms and financial intermediaries. Today, these values must also face new problems in the AI era. The EF Mandate notes that future centralization may not necessarily appear as a platform forcibly controlling you. It may also appear as a situation where you simply do not know how the system — especially AI — is making decisions on your behalf.

For example, when recommendation algorithms decide what content you see, when AI assistants filter information for you, when intelligent agents execute transactions on your behalf, and when identity, assets, and data are all abstracted behind interfaces, user sovereignty may be quietly diluted through one “more convenient” experience after another.

So Ethereum’s decentralization should not be understood only as node count, client diversity, or consensus mechanisms. It should be understood as a decentralized system in which no single entity can easily change the rules, users can verify system state, developers can build freely, applications can be publicly audited, and assets and identities do not need to be fully entrusted to platforms.

Vitalik has also emphasized that decentralization is not one feature of Ethereum, but the reason Ethereum exists. If Ethereum loses decentralization, it is merely a less efficient centralized service. This also explains why the Ethereum community has always emphasized that “everyone is a Builder.”

• In the Web2 era, most users are simply product users. Platforms define the rules, and users accept them. Platforms change the interface, and users adapt. Platforms shut down services, and users can only migrate or give up.
• But in the Ethereum ecosystem, builders are not only a small group of core developers. They also include wallet developers, DApp developers, node operators, researchers, educators, auditors, community contributors, and even ordinary users who carefully manage their private keys, learn on-chain security, and participate in governance discussions.

This means CROPS should not remain at the slogan level. It needs to be implemented through concrete products and actual user practices. Take a wallet like imToken as an example: security is not just a reminder message. It is a complete experience that includes mnemonic phrase management, risk alerts, DApp approval management, transaction parsing, phishing detection, and more.

Final Thoughts

Returning to Vitalik’s speech at the Hong Kong Web3 Festival, the surface-level topic was the technical roadmap for the next five years: scaling, zkEVM, post-quantum security, formal verification, privacy, block construction, account abstraction, zkVM, and more.

But at a deeper level, it was actually answering a question of values: when the entire industry is pursuing faster and cheaper systems, what should Ethereum optimize for?

The answer is not to reject performance, nor to reject user experience. The answer is to make on-chain applications easier for ordinary users to use — while ensuring that this improvement serves self-sovereignty, security, verifiability, and fair participation, rather than trading these values away for short-term efficiency.

The AI era will make this question even sharper.

Seen from this perspective, what truly matters in Ethereum’s next stage may not be whether it can become the fastest chain, but whether it can continue to serve as public infrastructure that is trustworthy, easy to verify, and minimally dependent on any single authority.