When AI Coding started taking over, why was the Ether Workshop still emphasizing safety and decentrization

Hong Kong Web3, Vitalik Buterin, without surprise, mentioned the old question: What is the Ether House going to be

In fact, in the past few years, the encryption industry has been used to discussing public-chain competition with performance indicators, such as who's higher TPS, who's faster in identification, who's lower in Gas, who's easier to understand as the next generation infrastructure...

But in 2026, whether it's the Ether Workshop or the new public chain, it's not just the internal issues of DeFi, NFT, L2 expansion and chain financeAI Coding Rapid Rainbow Inhalation, Formalization Validation and Zero Knowledge also show that progress is evident。

All of this means that the world facing the public chain is changing, so do we need a public infrastructure that can be validated by all, that can be withdrawn, that can be self-custody and that can be controlled without a single point

And how did the Ether House think and prepare for it

I. WHAT THE ETAL IS USED TO DO: THE OBSERVATIONS AND EQUIPMENT

"It is not about competing with high-frequency trading platforms, it is not about becoming the fastest chain, it is about becoming a safe chain, a decentrized chain, which will continue online, a chain on which you can always rely." In this speech, Vitalik re-explained the value of it with two very basic concepts: first, it is like a "public bulletin board"; and second, it provides "calculation" capability。

These two simple concepts, not only the new orientation of the Taifung that we have been elaborating in the recent past, but also the underlying reasons for the difference between the Taifung and ordinary Internet services, are almost general。

The so-called public bulletin board is not an abstract metaphor, but it means that the information can be published on the ETA, and everyone can see the content and sequence of the message - These messages can be transactions, Hash values, encrypted data, or other information that needs to be publicly committed, sequenced and validatedFrom "Global Computer / Clear Layer" to "Public Notice Board": What does Ether and Vitalik want to do)。

This is also the most fundamental difference between the Taifeng and the normal servers. Servers can be faster, cheaper and more efficient, but they usually require users to be confident that operators will not tamper with records, refuse services and shut down systems at critical times, as is the capability to be provided by the Taifeng。

COMPUTATION IS THE WAY IN WHICH PEOPLE ARE ALLOWED TO CREATE CODE-CONTROLLED SHARED DIGITAL OBJECTS, WHICH CAN BE ERC-20 TOKENS, NFT, ENS NAMES, OR DAO, CHAIN ORGANIZATION, FINANCIAL AGREEMENTS OR OTHER MORE COMPLEX APPLICATIONS。

SoBy 2026, it is easy to miss the problem that the ETA really is about to solve if it is still only "performance" to understand the ETA, or just TPS, Gas costs, confirmation speed。

The focus of the public chain expansion over the next few years has been not just on “making the chain faster” but on how to retain probabilities, decentrization and self-sovereignity in a more complex application scenario, in other words, not in order to transform the Ether factory into another centralised high performance system, but in order to allow more applications to operate without sacrificing the bottom confidence assumption。

This is also the key reason why Vitalik has re-examined L2 since the beginning of this year。

TO PUT IT SIMPLY, HE FELT THAT IN THE PAST THE MARKET HAD BECOME MORE ACCUSTOMED TO INTERPRETING L2 AS AN EXTENSION TOOL OF THE TAIFUNG, WHICH, WHEN THE MAIN NETWORK WAS EXPENSIVE AND BLOCKED, MOVED MORE TRANSACTIONS TO L2, BUT NOW THAT L2'S PHASED HISTORICAL MISSION HAS BEEN COMPLETED, IT MUST NOT REMAIN ON THE "SEGREGATION TRADE" LEVEL, BUT SHOULD BE A FRONT-LINE FOR THE EXPANSION OF ITS APPLICATION TO MORE APPLICATIONS。

This judgment is particularly important for today's Etherwood ecology。

In the past few years, the market has often reduced L2 to "a cheaper Etherwood". In the Vitalik framework, however, L2 is not simply a stratospheric layer, but rather a functional extension built around the public base of the Taifeng, L1 assumes the most critical commitments, settlements, data dissemination and validation roles, and L2 and the lower chain systems provide a higher frequency, more flexible and privacy-friendly implementation capacity around specific application scenarios。

This is why Vitalik is not willing to make the `fast' the primary goal of the EM。

Speed is of course important, but if the cost is that ordinary users are unable to run nodes, verify their state, and protect themselves from system problems, then the chain will gradually become a less efficient central service。

For the ETA, speed is a problem of experience, and safety and decentrization are the raison d'être。

IN THE ERA OF AI, THE VALUE OF THE ETHER HOUSE WILL BE RE-EXPANDED

One of the most interesting things about this talk is that Vitalik did not use AI as an external hotspot, but instead put it in the technical context of the future road map for the Etherak。

LIKE THE ETA COMMUNITY HAS BEGUN TO TRY TO APPLY AI TO GENERATE CODE PROOF THAT THE SOFTWARE VERSION RUNNING THE ETA HAS THE CHARACTERISTICS IT SHOULD HAVE. TWO YEARS AGO, THIS WAS HARD TO DO, BUT THE RAPID DEVELOPMENT OF AI IS MAKING IT EASIER TO SECURE SOFTWARE。

THE REAL PROBLEM BEHIND THIS IS THAT WHEN THE CHAIN OF BLOCKS IS CARRYING MORE AND MORE ASSETS, IDENTITIES, ORGANIZATIONS AND RULES, THE COST OF CODE LOOPHOLES WILL BE HIGHER. IF AI HELPS DEVELOPERS TO DETECT LOOPHOLES, GENERATE CERTIFICATES, AND ASSIST WITH FORMALIZATION, IT WILL NOT ONLY BE A TOOL FOR EFFICIENCY AT THE APPLICATION LEVEL, BUT WILL ALSO BE PART OF THE PROTOCOL SECURITY PROJECTWhen hackers "more efficient" use Al, Web3 's spear and shield' arms race, how does it escalate)。

HOWEVER, AI HAS A DEEPER IMPACT ON THE ETA, NOT ONLY AT THE DEVELOPMENT BUT ALSO AT THE USER LEVEL, PARTICULARLY BY CHANGING THE WAY ORDINARY USERS INTERACT WITH DIGITAL SYSTEMS。

As is well known, the human interaction of the past few decades has largely undergone several changes. The earliest users, through command lines and computer conversations, only a small number of people who know the technology can actually use complex systems, and then the graphical interface and the moving App is popular, and ordinary users can do it through buttons, pages and menus。

And nowAI IS MOVING INTERACTIVELY TOWARDS A NATURAL LANGUAGE, AND USERS NO LONGER NEED TO UNDERSTAND EACH STEP OF THE OPERATION, BUT SIMPLY TO SAY THE TARGET, THE SYSTEM MAY AUTOMATICALLY DISMANTLE THE PATH, CALL THE TOOL AND COMPLETE THE IMPLEMENTATION。

This change is placed in Web3 and will have greater impact。

To complete a cross-chain DeFi operation today, a user often needs to choose a network, confirm Gas, authorize a contract, execute a Swap, bridge assets, deposit an agreement, each step requires a signature, and each step can be wrong. If AI Agent becomes an important entry point for wallets and chain applications in the future, the user may simply say, "Turn a portion of my ETH into a stable currency and deposit it in a revenue agreement according to a low-risk strategy," and the rest of the route planning, protocol selection, trade simulation and execution may be done by smart agents。

This sounds like a significant lowering of the threshold of use, and, of course, the problem arises, after all, when users stop clicking on each step in person, and when AI Agent understands intent, calls contracts, initiates transactions on behalf of users, how can users confirm that it is not overstepping authority? How do you know that the path it chooses is not malicious? How can the ability to withdraw, validate and protect oneself be preserved without sacrificing the experience

This is the place where the value of the evaporation has been re-magnified。

AI CAN MAKE THE OPERATION MORE NATURAL, BUT NATURAL LANGUAGE DOES NOT BY ITSELF BRING TRUST, A SMARTER INTERFACE, IF THE BACK IS STILL AN UNVERIFIABLE BLACK BOX SYSTEMTHE USER SIMPLY WENT FROM "BELIEVING IN THE PLATFORM" TO "BELIEVING IN THE MODEL" AND THE ETHERPORT OFFERED MORE OF AN AI-ERA TRUST INFRASTRUCTURE。

To extend it, it actually makes the role of the wallet even more important. So the future wallet may no longer be just a "signature tool" or "assets list" but will gradually become the authority management between the user and AI Agent, the chain application, the identity system and the payment network, because the user needs to set the boundary through the wallet, for example, which operations are self-executable, which operations must be identified twice, which assets cannot be called and which authorizations need to be periodically checked and cleaned。

CROPS: FROM FOUNDATION CHARTER TO COMMUNITY CONTRACTS

Interestingly, just before Vitalik re-explained the Ether Workshop from the point of view of the protocol road map, the EF Mandate was also released by the EF Mandate Foundation, a formal confirmation of the route at the value level。

This Mandate file is presentedThe ultimate reason for its existence is to protect the self-sovereignity of users— Users are able to independently control their assets, identities, behaviours and choices without relying on any centralized intermediary。

With this in mind, EF Mandate proposed an abbreviation `CROPS', which is the initials of the Censorship Research, Open Source, Privacy, Security, which, in the presentation of the Foundation, must first retain these four attributes, without which the Taichat had no reason to be used, built or defended。

Objectively, the terms censorship, open source, privacy and security are not new in the Web3 context, and they have been discussed almost repeatedly since the beginning of the encryption industry, but they have changed dramatically today。

After all, in the early days of the encryption industry, these values were discussed more against centralized platforms and financial intermediaries, and now they are faced with the new problems of the AI era, as EF Mandate mentionedTHE CENTRALIZATION OF THE FUTURE DOES NOT NECESSARILY TAKE THE FORM OF A PLATFORM IMPOSING CONTROL OVER YOU, OR OF A SYSTEM THAT DOESN'T KNOW HOW TO MAKE DECISIONS FOR YOU。

FOR EXAMPLE, WHEN THE RECOMMENDED ALGORITHM DETERMINES WHAT YOU SEE, WHEN AI'S ASSISTANT FILTERS INFORMATION FOR YOU, WHEN SMART AGENTS EXECUTE TRANSACTIONS FOR YOU, WHEN IDENTITIES, ASSETS AND DATA ARE SEALED UP, THE SOVEREIGNTY OF THE USER MAY BE DILUTED IN A MORE CONVENIENT EXPERIENCE。

SoThe centralization of the Taifung is not simply understood as the number of nodes, customer diversity or consensus mechanisms, but rather as a decentrization system where no single subject can easily change the rulesAmong them, users can verify the status of the system, developers can freely build, applications can be audited publicly and assets and identities need not be fully entrusted to the platform。

Vitalik also emphasizes that decentrization is not a function of the Ether House, but rather a reason for its existence, because if it is lost, the Ether House is only a less efficient central service, which explains why the Etherdorf community has always emphasized that “everyone is Builder”。

• In the Web2 era, most users were only users of products. The platform defines rules, the user accepts rules; the platform changes the interface and the user adapts to it; the platform closes down the service and the user can only move or give up
• But in the ETA ecology, the builders are not just a small group of core developers, but also wallet developers, DApp developers, nodal operators, researchers, educators, auditors, community contributors and even every ordinary user who carefully manages his/her own private key, is safe in the learning chain and participates in governance discussions

This means that CROPS should not remain at the level of slogans, it needs to be achieved by specific products and specific behaviours, as in the case of purses like imToken, where security is not a hint, but a whole set of experiences such as assistive word management, risk alerts, DApp authorization management, trade analysis, fishing recognition, etc。

At the end

Returning to Vitalik's speech at the Hong Kong Web3 Carnival, it is ostensibly a technical route for the next five years: amplification, zkEVM, back quantum security, formalization, privacy, block construction, account abstraction, zkVM, etc。

But deeper, it's actually a question of valueAt a time when the whole industry is looking for faster and cheaper, what should the Ether House be perfected for

The answer is not to reject performance, nor to reject experience, but to make the chain more accessible to ordinary users, and must serve self-sovereign, secure, verifiable and equitable participation, rather than trade these for short-term efficiency。

THE AI ERA WILL MAKE THE PROBLEM EVEN MORE ACUTE。

From this point of view, the next phase of the ETA is really important, perhaps not as the fastest chain, but as the public infrastructure that continues to be the most trusted, easily validated and least dependent on single-point authority。