Xiao Feng's Discussion with Vitalik in Hong Kong

At ETHAsia during the 2025 Hong Kong Web3 Festival, Xiao Feng, Chairman of Wanxiang Blockchain, had an in-depth conversation with Vitalik Buterin, the founder of Ethereum. They reflected on Vitalik’s first visit to China in 2014. Vitalik emphasized the important role of Chinese miners, exchanges, and developers in the early development of Ethereum, shared his experience of learning Chinese, and acknowledged the rapid progress of the Chinese community in areas such as ZK (Zero Knowledge) and AI technologies.

When discussing application development, Vitalik stated that the foundation positions itself as an enabler rather than a leader and hopes that local communities will take the initiative to drive progress. He emphasized that the focus should currently be on DeFi (Decentralized Finance), RWA (Real-World Assets), and decentralized social networks, rather than just broadly discussing “applications.” He called for Chinese teams to participate in Ethereum’s POS 2.0 (Beam Chain) and other foundational research to drive protocol evolution.

On the topic of the impossible triangle — the challenge of balancing scalability, security, and decentralization — Vitalik noted that the layered architecture has already overcome about 70% of the technical bottlenecks. He believes that with the help of account abstraction, ZK technologies, and L1-L2 data sharing, it will be possible to achieve a WeChat-level Web3 application experience. He proposed that by moving service logic down to L1, the development threshold for L2 could be significantly lowered. L2 should also pay for the basic services provided by L1, enabling value to flow back.

Regarding the scope of “decentralization vs centralization,” Vitalik emphasized that real-world applications, such as RWA and identity systems, require some level of trust mechanisms. Blockchain should use cryptography and ZK to reduce trust costs, rather than aiming for extreme decentralization, thus improving fairness and efficiency.

On the global ecosystem, both parties agreed that developers from China and the U.S. form the two core forces. Vitalik pointed out that the foundation’s resources are limited and are primarily focused on underlying infrastructure and educational support. Commercial applications should be driven by the community itself. Finally, Xiao Feng suggested that the foundation set up an office in Hong Kong and restart Hackathons and Workshops in mainland China. Vitalik responded positively and expressed gratitude for the Chinese community’s support and contributions over the years.

The audio transcription was done by GPT and may contain errors. Please listen to the complete：

Xiaoyuzhou：https://www.xiaoyuzhoufm.com/episodes/67f707de623bc78c39a0b4fe

YouTube：https://youtu.be/Jnta4OLIAaw

Xiao Feng and Vitalik’s Reflection on the Early Days of Ethereum’s Adoption in China

Xiao Feng: Thank you, everyone. I want to share a secret first: 2025 marks the tenth anniversary of my first meeting with Vitalik. We met for the first time in Shanghai in 2015, so it’s quite a special moment for us to have this conversation again now. Of course, some of the feelings are personal, so we won’t go into detail about that in public. In 2015, Vitalik stayed in Shanghai for nearly three months. Although he had visited many times before, this was probably the longest stay.

So, I’m especially curious to know your impressions of China, Shanghai, the Ethereum community in China, and Chinese Ethereum developers from that time until now. This is a question that many Chinese developers are very eager to know.

Vitalik: Yes, I remember my first trip to China was in 2014. I traveled from Xi’an to Beijing, then to Shanghai, Hangzhou, Shenzhen, and a few other places. At that time, I could already sense that China had a lot of companies, teams, developers, and brands doing various things. I remember visiting exchanges like OKCoin and Huobi, and noticing that their staff numbers were larger than those of the biggest exchanges in the U.S. There were also many Bitcoin miners, and at that time, the industry was transitioning from FPGA to ASIC, around 2013 to 2014.

I recall one trip to Shenzhen where I visited a huge mining farm on the outskirts of the city. Those farms had a massive number of mining machines. At that point, I could already feel that China had a very powerful computing capacity, and various teams were actively pushing forward the development of Bitcoin and blockchain.

Back then, I noticed that these teams, projects, and companies in China were extremely large, but hardly anyone outside of China knew about them — there was virtually no recognition. At the international level, people mostly only knew about Ethereum and miners, but the Chinese ecosystem was already doing much deeper work.

I remember in 2015, you and some other companies were already doing very interesting things. From 2015 onwards, the Chinese community made many early and important contributions to the development of Ethereum, and we really appreciate your support starting so early.

I also remember that you bought a lot of Ethereum Foundation tokens. At that time, the Foundation really needed that money — it could be said that it survived because of that funding. Later, I also noticed that the Chinese community became more and more complex and diverse, with people starting to focus on many different directions, such as protocol research, cryptography research, algorithmic formulas, etc.

Actually, you could say that in 2015, aside from Israel and San Francisco, there weren’t many other particularly interesting blockchain projects around the world. But by 2016, I remember you were running many projects, and now we can see many excellent research teams working on ZK (Zero Knowledge), AI, and other areas. These all show that the quality of the community has become very high.

How Vitalik Learned Chinese in one Year and Delivered a Speech at Ethereum Events

Xiao Feng: The day before yesterday, I told Vitalik that we would be having this conversation in Chinese, haha. Just now, I was reminiscing about early 2015. At that time, when we were together, he couldn’t speak a word of Chinese. But one day, he told me that he was going to start learning Chinese. A year later, around August 2016, that September, Wanxiang Blockchain held a ‘Blockchain Week’ in Shanghai, and he was our main guest.

I contacted him in August and asked if he had a PowerPoint (PPT) presentation, so I could have it translated into both Chinese and English for the attendees to better understand. But he told me, ‘No need, I will write the PPT in Chinese and give the speech in Chinese.’ It only took him a year. I’m really curious — how did you manage to learn Chinese so quickly, not just speaking, but even writing a speech in Chinese? That’s much harder than just speaking.

Of course, I also recall in May 2016, when we went to New York, San Francisco, and London. I remember in the taxi in New York, Vitalik, myself, and Shen Bo from Distributed Capital were chatting in the car. At that time, Shen Bo and I were discussing something in Chinese, and before we finished, Vitalik took out his phone and asked me, ‘Is this what that word you just used means?’ That’s how he learned Chinese — just bit by bit. But I still want to hear how you managed to learn Chinese so quickly?

Vitalik: I’ve actually tried many different methods for learning a new language. My first step is usually using an audio course called Pimsleur. It has many language versions, and each course consists of 90 episodes, with each episode lasting 30 minutes. They start with the basics, such as the first lesson teaching things like ‘Hello,’ ‘I am in the U.S.,’ and ‘I am in China.’ Each lesson gradually becomes more complex.

The course method is repetitive listening, shadowing, and answering questions. Throughout the process, you keep speaking and repeating, gradually mastering basic expressions.。

Later, I added the ‘character recognition’ part because Chinese isn’t just about speaking — it’s also about understanding the concept of Chinese characters, which is more complex than many other languages. At that time, I used a flashcard app where I learned 20 new Chinese characters a day, and the system would repeatedly test whether I remembered their meanings and pronunciations. I kept practicing like this, slowly accumulating knowledge.

After that, the only truly effective way to learn was — communicating with people. Whether through voice or messaging apps, I took every opportunity to speak and write more. I remember when traveling in various cities in China, I would deliberately look at every street sign because they had both Chinese characters and pinyin. Through this constant exposure to everyday scenes, I was able to learn new things every day.

After about a year or two, the most important thing was to put myself in an environment where everything was in Chinese. That way, I could constantly understand more content and practice more expressions. When you’re in that environment every day, you naturally get more proficient.

The Contribution and Passion of Chinese Developers to Ethereum

Xiao Feng: I still remember in January 2016, Wanxiang Blockchain, together with Deloitte, held the first Ethereum Hackathon in Shanghai. Over 100 participants flew in from all over the world to attend, the majority of them from China, coming from places like Beijing or the southwest. But there were also some international participants, such as a high school student from Italy who specifically flew to Shanghai to attend. He came alone, and we formed a team for him on the spot. In the end, their team won first prize.

Later, this high school student reached out to us at Wanxiang Blockchain and asked if he could join our team. He said he didn’t plan to return to Italy. We advised him to at least finish high school first, then consider the future after getting into university. A year later, he went back to Europe and joined a project called LTA, where he later became a key technical member.

The reason I’m telling this story is to highlight that the developer community in China, since 2015, 2016, and 2017, has made tremendous contributions to Ethereum in terms of community building, technical development, and market promotion. They also have a very deep affection for Ethereum. This is not just about technical investment; it’s more about a profound passion.

I remember in 2016, when you gave a speech at the Wanxiang Blockchain forum, the reception you received was already that of a celebrity, even more enthusiastic than the cheers you get now. So, next, I’d like to ask, on behalf of the Chinese Ethereum community and Chinese developers — whether they are currently directly developing on Ethereum or not — everyone would love to hear your thoughts or those of the Ethereum Foundation. How do you hope to support Chinese developers and help them better play a role in the Ethereum ecosystem, build the community, and develop applications?

Vitalik: Yes, I think one important point is that we have already solved many of the information gap issues. In the past, communities in places like China, India, and Latin America often didn’t know about many things, but now this situation has improved greatly.

For example, during my three days here, I met a lot of community organizers who already have a deep understanding of topics like account abstraction, consensus algorithms, and so on. This shows that our information connection is getting better.

Additionally, we recently launched a project called Deep Funding, which is an open competition where anyone who can provide a good AI model to answer community questions can earn rewards. I met a participant last year at another event, and now they’re working on such a project. This kind of connection is becoming more common and bringing a lot of positive effects.

As for what we need most right now, it’s actually more specific ‘applications.’ But sometimes I don’t really like the word ‘application’ because it’s too broad. For example, Facebook is also called an application, Tencent is also called an application, so if we just use ‘applications’ to describe Ethereum, it’s hard to capture the unique value of this industry.

I’d rather see more people pushing deeply in some specific areas, such as DeFi (including RWA and other real-world assets), decentralized social networks, info finance, and so on. It’s not about having a hundred people working on one application, but rather spreading out across different directions, with twenty to thirty people deeply pushing in each direction. That way, the community becomes more interesting and more dynamic.

In fact, there are already many teams working on community building, discussing topics, organizing events, and attracting developers to participate. These are all very encouraging things. The Ethereum Foundation also has many local community support programs, like the various activities you’ve seen these days. We provide support in different forms — such as funding, tools, platform resources, and so on.

Our foundation’s goal is not to control the Ethereum ecosystem, but to foster its self-growth. The ideal situation for us is that the things we do today will no longer be necessary in the future because the ecosystem will have grown strong enough and be sufficiently self-governed.

Another important point is about Layer 1 (L1) development. Actually, anyone can start working on it by themselves. You don’t need to know anyone in particular, nor do you need approval from the foundation. As long as you have the passion and ability, you can directly participate.

Recently, we’ve been working on solving another issue, which is defining and advancing the decentralized standards for Layer 2 (L2) — stage 0, stage 1, and stage 2. Right now, many L2 projects claim to be decentralized, but the standards are unclear and are mostly marketing tactics. We want to establish a clear technical standard, so that if a project meets certain security requirements and technical specifications, it can be recognized as a true stage 2 decentralized network.

This approach can make the entire ecosystem fairer and make it easier for newcomers and small teams to get involved.

Our approach to L1 development is the same. Currently, our researcher Justin Drake is leading a project called Beam Chain, which can be seen as an attempt at Ethereum’s PoS 2.0. Over the past decade, we’ve done a lot of research on PoS (Proof of Stake), and now we have a deeper understanding of algorithms, quantum resistance, and game theory. We hope to consolidate these findings into a new L1 experimental project that allows global teams to participate.

At this point, there are at least 7 candidate teams for Beam Chain. I even know that teams from institutions like Shanghai Jiao Tong University and other universities in China are conducting related research. If anyone is interested in this direction, I strongly recommend that you follow it and get involved.

The issue we face with L1 research right now is that if you don’t know anyone in the relevant groups, even if you have strong technical skills, it’s very difficult to participate. This is a problem we are particularly focused on solving. So, the development of Beam Chain will be completely open, attracting more new teams to start from scratch and contribute to building the foundation layer of Ethereum for long-term evolution.

What Are the Key Milestones for Achieving ‘WeChat-Level’ Blockchain Applications?

Xiao Feng: Haha, yes. After hearing what you just said, I actually have three questions I want to discuss with you. The first question is, as we all know, there is a theory called the Blockchain Trilemma — decentralization, security, and performance cannot all be achieved simultaneously. You can only balance two of these, sacrificing the third. This is not just a blockchain issue; many other systems, such as exchange rate mechanisms, also have similar logic. Essentially, it’s a mathematical problem.

So, how can blockchain significantly improve performance and scalability while ensuring decentralization and security? If efficiency and scalability cannot break through, large-scale applications cannot be supported. From the perspectives of cost and user experience, it seems that solving this trilemma on a flat level is not feasible.

My personal view is that, to this day, I still believe the layered architecture proposed by Ethereum — using the L1 and L2 model — is the best solution to solve this ‘impossible trilemma.’ You also mentioned the relationship between L1 and L2.

But now, I want to ask from a different perspective — how do we solve the issue of ‘usability’? That is, how do we lower the barriers to using this technology, reduce costs, and enhance the user experience to the highest level? Because only when a high-tech product or service is truly easy to use, can it be widely adopted by the masses.

Our industry has always been asking one question: When will the ‘killer application’ in the blockchain field appear? That is, the application with a large user base. The premise of large-scale applications is higher performance, lower costs, and stronger scalability.

Just like how computer systems evolved from the original DOS command-line operating system to graphical interfaces, and then to mobile internet apps, each step lowered the barriers to use. If we were still using DOS command lines today, it’s hard to imagine that over a billion people would be using computers. So, after Microsoft introduced graphical operating systems, the number of users expanded from millions to billions, and eventually, the popularity of mobile internet allowed users to perform operations with just a tap, which led to today’s global user base of over 5 billion internet users.

So, my question is: While our current layered architecture is good, the barrier to ‘usability’ is still too high. When will the ‘app moment’ for blockchain happen? In other words, a complex Web3 system can be accessed by ordinary users through a super app without them realizing it, just like WeChat — no matter how complex the system, users can just tap an icon to use it. When will we see this ‘WeChat-level’ blockchain experience? I think that will be the real turning point for the explosion of killer applications. What do you think?

Vitalik：Yes, I think it will still take a lot of time to reach that moment. We are indeed making progress, but there are still some technical issues that haven’t been fully resolved.

As I mentioned in my talk this morning, Ethereum is still facing scalability challenges. You may remember that in my 2015 talk, I mentioned four technical challenges that blockchain development faces: the first is scalability, the second is PoS (Proof of Stake), the third is privacy protection, and the fourth is security.

Now, Zero Knowledge (ZK) technology has developed rapidly, and scalability has improved significantly. The PoS mechanism has successfully gone live, but there’s still room for further optimization. As for security, we are not only focused on the security of the protocol itself but also on the security of user accounts.

For example, there have historically been two extreme options for managing assets on-chain: one is complete self-custody, where you store private keys on your computer, phone, or even on paper. This is very decentralized, but it is too complex and too risky for the vast majority of people. Even for me, it’s difficult.

The other option is using centralized services, but this brings trust issues, as seen in the collapses of MT.Gox and FTX.

So, our current focus is on finding a better balance between these two extremes. For example, we now have multi-sig, account abstraction, ZK Email, and other solutions aimed at improving user experience and reducing security risks.

Returning to what you mentioned about the “app moment,” I think from the four goals we set in 2015, today we’ve probably achieved 50% to 70%. As long as we solve the remaining technical challenges, applications will definitely come.

Why do I believe this? Because we’ve already seen some cases, such as Worldcoin now having 10 million users, and institutions like Sony, Kraken, and Deutsche Bank deploying L2 on Ethereum. There are also many countries, like Argentina, Turkey, and regions in Southeast Asia, where large numbers of users are using stablecoins, ETH, and DeFi services.

These all show that the infrastructure is nearly ready, and as long as we can break through the final user experience bottlenecks and developers are willing to invest, killer applications are entirely possible.

So, I’m still optimistic. We will eventually reach that “app moment.” OK.

How Can We Make L2 Lighter and Easier to Build by Shifting ‘Dirty Work’ to L1? L2 Pay for L1 Services.

Xiao Feng: You just mentioned that traditional large institutions like Deutsche Bank have already started working on Layer 2 (L2) development on Ethereum. Last December, we also launched Hashtag Chain, an L2 based on Ethereum. But during the process of building HushKey Chain, I’ve always had a feeling — that compared to developing an app, the cost of building an L2 chain is still very high.

For example, if you want to build an L2 from scratch, it might take 20 engineers and a whole year of work. You not only need to build the chain itself, but also the browser, wallet, development tools, etc. It’s far from as simple as “writing a chain.” And the technical operations cost alone for maintaining the chain could reach millions of dollars annually.

Of course, I understand that L2 cannot be as cheap, fast, and low-cost as mobile apps, but if you put yourself in the shoes of those who want to start an L2 project from scratch, the barrier is really high. They need to hire 20 professional developers and solve all kinds of integration and operational issues.

So, I have an “outlandish” suggestion: could we push as many of these “dirty and difficult tasks” as possible down to L1? In other words, could L1 provide more basic capabilities to make L2 itself lighter, thinner, and easier to develop? This way, you wouldn’t need to invest so many people, time, and resources into creating an L2 like HushKey. Do you think this is possible? When do you think it can be achieved?

Vitalik: We are actually already working on this. I can provide some concrete examples to illustrate.

One of the high costs of L2 today comes from integration. In other words, developers building applications on L2 typically need to connect to various basic services, such as Oracles, Filecoin, identity systems, etc., all of which require separate deployment and management.

What we are working on now is making the connection between L2 and L1 more efficient. I mentioned this topic in my talk this morning. For example:

The first step is improving the communication efficiency between L1 and L2. Currently, it takes about a week to submit state back from L2 to L1. We hope to shorten this to one hour initially, and ultimately to 12 seconds. This would significantly reduce data transfer latency.

The second is introducing an opcode in L1 called L1SLOAD, which means that the EVM on L2 can directly read data from L1. For example, if each L2 needs to independently deploy an Oracle, it’s quite a waste of resources. But if you can directly read the Oracle information from L1, you can share data and reduce redundant deployments.

Another example is wallets. If we want users to support Account Abstraction, such as allowing users to freely change private keys or switch signature algorithms, the current method is to send a transaction on each L2 to update the settings every time, which is cumbersome. But if we deploy these key account information on L1, we only need to make a change on L1, and each L2 can sync the state. This would solve the problem efficiently.

This is also the concept I proposed with the “Keystore Wallet” — the core wallet state of the user is stored on L1, and all L2s can directly read it, truly achieving a shared account structure.

So, as long as we can design efficient interfaces between L1 and L2, we can make L2 lighter, and developers won’t have to reinvent the wheel.

Moreover, this model can bring an additional benefit: we can drive the emergence of “hybrid apps” (L1 and L2 mixed applications). These applications will no longer be exclusive to L1 or L2, but will be able to use data and services from both layers, enabling a more powerful combination of functions.

Xiao Feng: If L1 can take on more of the “dirty, difficult, and tedious work,” that would indeed address a core issue that many people are criticizing now. Many say that L2 creates a lot of value, but that value is not fed back to L1. L1 provides the infrastructure but ends up not capturing the value it deserves.

If L1 can provide more foundational services to make L2 lighter and easier to use, then L2 should pay for these services, returning some of the value to L1. This is not only technically reasonable, but it also forms a “value feedback” mechanism in the economic model. Achieving this would be great news for the entire Ethereum ecosystem.

You also mentioned decentralization earlier. It is true that some projects are decentralizing just for the sake of decentralization, putting up a photo and claiming to be “Ethereum native” or “decentralized.” But when it comes to the application layer, the issue becomes much more complicated.

For example, as you mentioned with RWA (Real World Assets), putting real-world assets on-chain is inherently impossible to fully decentralize. Because any issuance of an RWA is legally considered a securities issuance, it inevitably involves issuers, operators, approval processes, and regulatory bodies, which will inherently contain centralized elements.

So, I think we need to be more realistic. At the infrastructure layer, Ethereum must remain decentralized, open, and permissionless. But when it comes to the application layer, decentralization is not a one-size-fits-all solution. The key is finding a balance between “minimizing trust” and “maximizing efficiency,” rather than sticking rigidly to a particular form.

How Do We Balance Decentralization and Centralization in Blockchain Solutions?

Xiao Feng: The issuance of a securities-like asset naturally involves the issuer, operator, approval parties, and regulators, which will inevitably contain some centralized elements or a certain degree of centralization. If we want to integrate blockchain with traditional finance, using Ethereum or other blockchain systems to create new financial infrastructure, we must face this reality.

The reason finance is strictly regulated is because it has strong negative externalities. In traditional finance, behaviors such as “cutting the leeks” (fraudulent practices) are illegal because they reduce social welfare, create risks, and generate unfairness, which is why regulation is needed. And this negative externality cannot be ignored once it is introduced into the blockchain world.

As blockchain technology increasingly integrates with the real world, I believe it makes sense to “layer” the system. As infrastructure, Ethereum L1 must adhere to decentralization. Just like the internet’s foundational protocols (such as the IP protocol) are open-source, permissionless, and anyone can build a network using the IP protocol without needing to obtain permission from others. This openness is the foundation for the development of the internet.

Blockchain technology is the same; as a foundational protocol, its core value lies in being open-source, permissionless, and decentralized. But when it comes to the application layer, the situation changes. Internet applications are essentially centralized. The Web3 world doesn’t necessarily have to replicate this structure, but it also cannot ignore the rationality of centralization in certain areas. How do you view this balance between “centralization vs decentralization,” especially at the application layer?

Vitalik: Yes, at the application layer, if we are dealing with real-world economic activities or interpersonal relationships, complete decentralization or complete trustlessness is not feasible.

Why do I say this? Because many real-world actions cannot be proven through cryptography. For example, if I want to sell you a phone, you send me 0.5 ETH via Ethereum, and then I send you the phone through DHL. After you receive the phone, you confirm the transaction on-chain.

But the issue is that during this process, the blockchain cannot verify whether I actually sent the phone. You can’t use cryptographic methods to prove “I put the phone in the package” or “DHL really picked up the phone.”

Perhaps in the future, we could use advanced Zero Knowledge (ZK) technologies, such as proving that I really handed the phone over to DHL, or DHL could provide a verifiable on-chain proof of delivery (such as ZK + TLS Notary) — but even so, you would still need to trust DHL and believe that I really put the phone inside. These kinds of problems are widespread in the real world and are areas that cryptography cannot fully cover.

Of course, the role of blockchain and cryptography is to reduce this need for “trust” as much as possible. When the trust threshold is very high, only widely trusted large institutions can participate, excluding ordinary people, which is not fair.

Through blockchain technology, what we can do is enable more people to have the right to participate. For example, RWA (Real World Assets) can be issued in the form of tokens. For instance, I issue a Filecoin token with a total of 1 million units. I only issue 1 million on-chain, and the system can publicly verify that no extra tokens have been issued. Through ZK technology, it can also be proven on-chain that I actually have USD or HKD assets in a bank account corresponding to these 1 million tokens.

Another example is when a government issues digital identity IDs. The public might worry that it secretly issues “ghost IDs” to manipulate elections or social platforms. If these IDs are issued on-chain, with the total amount visible, the public can confirm that the government has only issued 1 million IDs, not 2 million. This mechanism can greatly reduce the trust cost in the government.

This mechanism of reducing trust can, on one hand, improve efficiency, and on the other hand, enhance fairness. Because a high trust threshold means that small organizations or individuals struggle to participate, and only large institutions with widespread trust can enter. However, once verification mechanisms can replace trust, more people will have the opportunity to participate, making the economy and society more open and fair.

Of course, in certain rare scenarios, such as the verification of mathematical competition problems, complete decentralization can indeed be achieved. For example, I design a smart contract, and if you can submit a valid ZKP (Zero Knowledge Proof) that satisfies a certain formula, I will reward you; otherwise, I won’t. These types of scenarios are suitable for fully automated, verifiable logical loops.

But in 99% of real-world economic activities, such an idealized, fully decentralized mechanism is not feasible.

Xiao Feng: Yes, I completely agree. This is actually something we must constantly remind ourselves of when developing Web3 or blockchain applications — not everything can be solved with decentralization.

Of course, for the parts that can be solved with decentralization, we should use it as much as possible because it can reduce trust costs and improve system fairness, efficiency, and transparency. But for scenarios that require connection to the physical world and behavioral verification, such as phone transactions, physical logistics, and so on, on-chain mechanisms currently cannot fully cover these areas.

The example you mentioned is a good one. We still have to trust DHL to deliver the phone, and we must believe that they actually completed the delivery.

So, we should not be fixated on “100% decentralization from the bottom layer to the application layer.” Decentralization is not the goal; it is simply a means to achieve a more efficient, fairer, and lower-cost economic system. We should make good use of it, but we also need to accept its boundaries and limitations.

And now, with Ethereum entering the layered architecture phase, I believe this is the prerequisite for blockchain to move toward large-scale applications.

How Can Chinese and American Developers Collaborate to Launch Large-Scale Blockchain Applications?

Xiao Feng: Hmm, this new phase has already arrived. As Vitalik just mentioned, many key technologies, including the continuous optimization of L2, are advancing rapidly. We are also gradually approaching the threshold for large-scale blockchain applications.

From the internet, AI to cryptocurrencies, we can observe an interesting phenomenon: among the top 15 largest internet platforms globally, aside from the United States, most are Chinese companies. The same is true for AI large models; besides the United States, China has also proven its R&D capabilities, such as through DeepSync, Huawei, Baidu, and others. However, countries like Japan, India, the European Union, and the UK have not yet put forward competitive large models. In other words, technological capabilities are not evenly distributed worldwide.

I believe blockchain will follow a similar pattern. While developers are spread globally, they are mainly concentrated in two language communities: the English-speaking world (centered around the U.S.) and the Chinese-speaking world (centered around China). The scale of English-speaking developers is indeed larger, but Chinese-speaking developers also form another massive ecosystem.

So, here’s the question: If we are truly aiming to push Ethereum into the stage of large-scale applications, how should Chinese developers, product managers, operators, and users participate? In fact, Chinese people have very strong capabilities at the application layer. Since the internet era, they have been excelling in competition, producing a large number of excellent product managers and platform operators.

For example, more than half of the top ten digital asset exchanges in the world are founded or operated by Chinese people. Why? Because the technical nature of Bitcoin is the same across all exchanges, but customer experience and user service are areas where Chinese people clearly excel.

So, if we recognize that Ethereum is entering a development stage focused on applications, how can we better leverage the strengths of Chinese-speaking developers? Not only developers but also operators and users. The Chinese-speaking internet user base already exceeds 1.4 billion, and with the addition of Hong Kong, Macau, Taiwan, and overseas Chinese, it’s a massive market. Do we have a strategy for this group to incentivize them to develop applications on Ethereum and drive adoption?

Vitalik: Yes, this is a very key issue.

First of all, I believe that the development of applications should not be “directly handled” by the Ethereum Foundation. We don’t want, five or ten years from now, everyone still relying on the foundation to drive ecosystem development. We would prefer that every local community around the world can grow and have the ability to independently drive projects, organize, incubate, and implement various initiatives.

In recent years, we’ve noticed that developer communities in China, Asia, and even Africa have been growing, and they have established increasingly more communication with our foundation, especially in terms of promoting local developer organizations, project incubation, and talent cultivation.

There are two reasons why we don’t want to do everything:

The first reason is to avoid centralization. A healthy ecosystem cannot rely entirely on a centralized institution. We want Ethereum to evolve on its own, not for the foundation to lead every application or project.

The second reason is resource limitations. You might not believe it, but the amount of ETH currently held by the foundation is even less than what Justin Sun holds (laughs). Our budget is very limited, and many organizations in the community actually have more financial power than we do.

Moreover, fundamentally speaking, applications can be self-sustaining. If you create a valuable application, it can attract users, gain investment, and become profitable. Therefore, we focus more on areas where “if the foundation doesn’t do it, no one else will.”

For example, core R&D of L1, ZK Email, Account Abstraction, ZK Wrapper, and other foundational infrastructure components. These are high-barrier, low-commercial-return areas, so we are willing to provide more support.

But for DApps, wallets, and other application development, they are already profitable and no longer require the foundation to drive them directly.

So, our strategy is to focus on supporting the growth of local communities. We hope to help more developers understand Ethereum’s core concepts and architecture, such as L1, L2, ZK, clients, etc., and guide them to explore specific use cases.

We cannot establish and manage all the communities around the world, so we need to collaborate more with local organizations, helping them build their own developer ecosystems in their cultural and linguistic contexts.

In this regard, we have already made progress. Compared to two years ago, communication and collaboration with developers in Asia are much smoother.

We also encourage people to express their needs proactively. If you encounter problems or have specific needs while working on a project, our foundation is also willing to provide support. We hope that within the next year or two, every city will have a stronger developer community, nurturing the next generation of core Ethereum builders.

This transformation has already begun, and we believe it can grow better and broader.

Xiao Feng’s Proposal for Ethereum to Set Up an Office in Hong Kong and Restart Hackathons in Mainland China

Xiao Feng: Yes, as you just said, it’s correct for the foundation to focus on those things that application developers cannot complete independently, such as ZK (Zero Knowledge Proof), Layer 1 protocol research, and so on. Because the foundation should have a clear focus. When application developers build on Ethereum, they may indeed need the foundation and Ethereum community’s support to understand how to do things better on Ethereum. This support doesn’t necessarily have to be financial; it is more likely to be in the form of technical guidance, resource matching, or even exposure channels.

So, the reason we’re hosting this meeting at ETH Action today is to invite Ethereum community representatives from various countries in Asia, and we hope to have such an annual gathering of the Ethereum community in Asia. This will allow everyone to exchange ideas on how to promote Ethereum applications and technologies in Asia. This itself is a great platform.

On this basis, I have two suggestions for the Ethereum Foundation:

The first suggestion is that I hope the Ethereum Foundation can set up an office in Hong Kong. This would be extremely helpful for the entire Ethereum community in Asia, including application developers. This office wouldn’t just be a symbolic presence but would serve as a contact window that can provide practical technical consultation, guidance resources, and exposure opportunities. Many developers may not necessarily need funding from the foundation, but they certainly need a place to seek support and make connections, especially for some startup projects.

In fact, one of the purposes of our meeting these past two days is to help these developers and projects gain exposure, build connections, and find collaborative resources. This is the first suggestion, and as you can see, everyone has already shown their support with applause, haha.

The second suggestion is that I think the foundation, including yourself, should return to mainland China and restart face-to-face interactions with Chinese developers. You’ve done a great job with this in the past, and I remember that in 2017 and 2018, you frequently came to Shanghai. This interaction was interrupted later due to the pandemic.

Now that the pandemic is over, it’s a great opportunity to resume this kind of interaction. For example, the foundation could consider restarting hackathons and workshops in mainland China; you could personally visit various places and engage in face-to-face discussions with developers. This October, we will hold the 11th Blockchain Global Summit in Shanghai. You have already participated in the previous seven editions, and we would be very happy to have you join again.

At Wanxiang Blockchain, we are very willing to assist and promote this initiative. We hope to restart Ethereum workshops and hackathons in multiple cities across mainland China. This format is not just an event, but a substantive form of guidance for developers, helping them master Ethereum technology and build more practical, real-world applications. These are the two suggestions I wanted to bring to the foundation and to you today.

Vitalik: Well, thank you very much, Mr. Xiao. Thank you for all the support you’ve given to the Ethereum community, the foundation, and the entire blockchain ecosystem over the years. I really appreciate it.

Xiao Feng: Well, thank you, everyone. Our discussion ends here today. I can guarantee that we didn’t rehearse or communicate beforehand.

Vitalik: Hahaha, OK, thank you.