This thing, we can also do it, it is safer and feasible, such as quadratic fundraising or various other Plurality mechanisms and implementations, so these two layers are consistent, and both of us need them.
(End of Chinese interview)
Here is the translated English version of the interview transcript:
Dong Qu: Vitalik, as a supporter of cryptocurrency and a cypherpunk, Ethereum embraces and emphasizes the values of cypherpunk, such as decentralization, transparency, and privacy. So, many people know about Ethereum but are not familiar with Plurality.
Can you please tell us how people can connect Ethereum with Plurality? Also, can you tell us why you like cypherpunk and how you define cypherpunk? What role does Plurality play in cypherpunk?
Vitalik: I think the idea of cypherpunk is fundamentally about freedom and openness, creating something that is as participatory, accessible, open, and fair as possible, and protecting people’s rights and related matters. The method is to adjust and leverage digital technology to achieve these goals.
I think this has always been the dream of people trying to realize the cypherpunk ideals. It has almost always been like this. And I think this has been the dream of people for decades, trying to create things like digital cash, digital private property rights systems, and digital voting systems, and doing all these things in a secure way through cryptography instead of relying on trusted centralized authorities.
Because once your security depends on a centralized authority, that centralized authority can easily abuse that power. And in the physical world, of course, sometimes it is difficult to avoid central authorities because in the physical world, attacks are often easy and defense is difficult, which still causes a lot of problems. But if something like this happens in the digital world, the situation is just the opposite.
Vitalik: You can have things like cryptographic keys, you can sign, and even the most powerful attackers cannot break these signatures. The more things we do in the digital world, such as building various digital technologies and systems, the less we rely on a centralized authority, which reduces the risk of being affected by the problems of these centralized institutions. What we want to explore is how far we can go in this regard. This idea began with the digital cash system invented by David Chaum in the 1980s.
Many things like this, such as anonymous email systems and mix networks in the 1990s. As a result, we saw that Bitcoin created the first completely decentralized digital currency. I think now in 2024, we have a lot of different tools to build all these structures, and we have seen that we can build decentralized currencies, we can build voting systems, we can build various different things.
At the same time, we need an organized movement to actually implement these things. If there is not a conscious and focused movement trying to do these things, people in the crypto field may just try to make money by speculating on cryptocurrencies.
Therefore, we see that people outside the crypto field are trying to collect data and just trying to make money by selling advertisements and getting people to buy useless products. We actually need to create this organized movement, or at least try to create an organized software tool that follows these principles, collaborates well with each other, and shows people what a better, more open, more secure, and more privacy-focused software world could be like.
Dong Qu: With recent efforts to develop more public goods and lay the foundation for integrated civic technology, Ethereum aims to achieve “high-resolution democracy.” So, what is the difference between Plurality’s digital democracy and Ethereum’s high-resolution democracy?
Vitalik: Yes, when I think of high-resolution democracy, my idea is that because we have tools like modern information technology, we can extract richer and more detailed information about people’s preferences than 20 or 30 years ago. Traditionally, we only hold elections every four years, which basically gives us one or maybe two bits of information every four years, averaging less than one bit per year. This is a very low-bandwidth communication method.
But of course, if you only have tools like paper, or systems like 200 years or even 50 years ago, that may be the best you can do. But now, we actually have all these very powerful information technologies. The question is whether we can use these technologies to actually obtain more information about people’s preferences and achieve it through these mechanisms.
In our modern internet, we have many examples of what I call “micro-democracy.” So, take Twitter, for example. You know, you can post a tweet, and people can like it. People can retweet it. You can see how many likes it has. On Reddit, you can see upvotes, you can see downvotes. These are forms of micro-democracy. It’s like a real-time referendum on whether people think what this person said is good or bad, and these referendums happen millions of times every day.
And their results affect who sees what content. So, we have a lot of these examples of micro-democracy, and we also have macro-democracy. But at the same time, there are two challenges, namely “macro-democracy is inefficient, and micro-democracy is insecure.”
In the case of social media, there are many fake accounts, and voting in micro-democracy is not anonymous. So, you know, if you like a tweet, there are many ways it can be traced back to you. And if you live in the wrong country, you might even be punished for it. So, these are the challenges of majority rule.
We need to see if we can create a voting mechanism that is both micro and macro, secure, and good at extracting large and accurate types of information. At the same time, we need to do it in a way that has never been tried before. I think one major success we’ve seen in the past two years on social media is the emergence of community annotations.
This is a new algorithm on Twitter that can identify viewpoints that everyone agrees on, even if these people have different opinions on other issues. It’s mainly used to show clearer information or background information about various tweets. This approach is particularly important for political content because on the one hand, there are concerns about fake news, and on the other hand, concerns about being suppressed by dictators or small circles.
So, this more democratic mechanism that can avoid being controlled by a minority or even try to bridge different groups. Its goal is to avoid being completely dominated by a certain group, and that is a very powerful feature that many people are looking forward to.
So, our goal is to find at least 10 to 20 examples in various fields, such as social media, funding decisions, local governments, or national governments, that conform to this spirit. In addition, companies like Apple have many applications in the blockchain field and have begun to implement them.
Upgrading Infrastructure Trust
Dong Qu: Okay, now let’s talk about the second part, which is trust in infrastructure. Glenn, Audrey, and Pluralist all implied that the most important infrastructure is the power distribution of Plurality. For Plurality, its goal is to be popularized by non-governmental organizations and national governments, not just national governments. So, how does Ethereum plan to convince these opponents to trust and integrate with Ethereum?
Vitalik: Yes. Institutions like Ethereum and blockchain technology have been something people have been trying to do for a long time. Obviously, so far, it hasn’t been successful. I remember five to ten years ago, the exciting trend was private blockchains, right? Or sometimes it was called consortium blockchains, private chains, enterprise chains.
It’s the idea that you can try to create something that is more decentralized than a server, but not as fully decentralized and scary as a public blockchain, and try to find a compromise in the middle. But this compromise has obviously ultimately failed, just like we haven’t seen many enterprise blockchain applications really work.
Vitalik: I think the reason is that they ultimately created the wrong type of compromise because by not working with public chains, they basically gave up the ability to benefit from the network effects of a truly large community. Personally, I noticed a phenomenon when I tried to talk to many of these teams. They would create a consortium first, and then the first five customers would appear, and they would be very happy and excited, but then it would never grow beyond five. Because if you are one of the first five, then you have a lot of control, and you can set the direction. But if you are the sixth or thirteenth, then you know, you don’t have power, and you are joining a system controlled by these other groups.
I think a better way of finding a compromise between centralization and decentralization is to combine blockchain with zero-knowledge proofs. You can use blockchain, not necessarily to publicly expose all data, but to expose the hash values of the data. Then, you use zero-knowledge proofs and publish the proofs together with each hash value to prove that the updates to the hash values are correct, regardless of how they are updated.
The interesting thing about this approach is that although it is not directly related to the blockchain from a technical point of view, it still brings many benefits. Users can still have Merkle proofs and can actually prove that the entries in their account balances are correctly computed.
Vitalik: You see, I think one of the initial, simplest, and most interesting applications is providing funds proof to exchanges in the short term. Exchanges have always faced a big problem: what if they are only partially reserved and are stealing people’s money?
Like Mt. Gox in the past decade and FTX in this decade, if no changes are made, similar problems may occur in the next decade. With funds proof, you can create a proof that shows for each coin in users’ assets, in their account balances, you actually have a corresponding coin, and these coins are stored somewhere under your control.
Interestingly, you can describe this as a funds proof application based on zk-SNARKs, or you can describe it as validity, thereby transforming exchanges into a kind of second layer of Ethereum. Technically, both explanations are valid. This kind of institutional application of Ethereum now seems very meaningful, and it’s just the beginning.
I believe the benefit of this method is that you can basically maintain the original operation of the application. For example, if you have a server, you can keep that server and just add additional software next to it. This software reads the database, performs some additional calculations, generates proofs, and then publishes them to the chain. By doing this, you are adding additional security to the existing system and giving users additional capabilities, making the exchange more secure. Therefore, you can do this and then use these proofs, which can be used later. So I think one of the reasons why this hasn’t been implemented yet is actually a technical capability issue, with much of it relying on zero-knowledge proofs.
Blockchain provides guarantees such as executability, transparency, and resistance to censorship, but they come at the cost of scalability and privacy. Scalability and privacy are both very important. What do zero-knowledge proofs do? They allow you to regain scalability and privacy. Five years ago, zero-knowledge proofs were mostly theoretical academic things. Now, privacy-oriented cryptocurrencies like Zcash, which are based on zero-knowledge proofs, have been launched at the end of 2016.
Vitalik: I remember in the initial version, to generate a proof, you needed a laptop and it was impossible to do it on a mobile phone. Even on a laptop, generating a proof took 90 seconds. But today, you can generate a proof on a mobile phone in less than four seconds, so this is the user experience.
Then there’s the developer experience. To create a zero-knowledge proof application before 2021, you basically had to be a cryptographer. But today, you only need to write code with Sercom, and it will be compiled. You don’t actually need to personally understand or be involved in cryptography. So everything is becoming more powerful, both in terms of user tools and developer tools.
So from a technical perspective, many issues are actually being solved. So now I think it’s time to really start identifying which specific things really make sense and start implementing them.
Dao Qu: Ethereum still has high transaction fees and low transaction speeds. Vitalik, do you also think this is the main reason why the Cypherpunk spirit cannot be widely adopted? And to what extent can the upcoming London upgrade solve this?
Vitalik: Yes, I think high transaction fees can indeed be a major factor hindering the success of many interesting applications in Ethereum. The main way the Ethereum ecosystem addresses this problem is through Layer 2 solutions, and Layer 2 is already starting to exist.
I think a major achievement last year is that we have at least one or two ways, like Arbitrum, that have achieved what we call Phase 1, which basically is a phase where it actually delegates most of the power to a real proof system. This system directly reads and verifies the computation of the Ethereum Virtual Machine (EVM), so it is actually more secure than just putting all the money into a multi-signature wallet. So Layer 2 is making great progress, and there will be more progress this year.
However, at the same time, Layer 2 has a scalability bottleneck, which is that they still need to put a certain amount of data on the chain for each transaction, right? EIP-4844 (also known as the Blobs proposal) aims to create additional space where more data can be put on the chain and processed in a cheaper way.
In the short term, this basically means processing in a way that makes it easier to download in parallel and easier to discard after a period of time. But in the long run, there is a goal to use a technique called data availability sampling, which allows nodes on the chain to verify the existence of data without having to download the entire data. So the goal of all these things is to increase the data space that Roll-ups can use, thereby reducing transaction fees and further improving the scalability of Roll-ups.
Dao Qu: Now we move on to the third part, the experience of technological optimism. We see that you mentioned d/acc in your blog. While technology can accelerate social progress, we cannot judge whether its direction is good or bad. The Ethereum merge may be the biggest success in blockchain history as it transitions from Proof of Work (PoW) to Proof of Stake (PoS).
As a leader in this transition and achieving success in this merge, what compromises and difficult decisions did you and the community make? And are there any experiences and insights from open collaboration that have played a crucial role in ensuring the direction is correct?
Vitalik: I think one important thing we did regarding the transition to Proof of Stake (PoS) is that it was a topic almost discussed in the Ethereum community from the very beginning. I think when you make a major change to a system that many people rely on and have expectations for, it is very important to clearly communicate what you are doing to the community and respect people.
An example I mentioned in that blog is the recent incident at OpenAI. The board tried to fire Sam Altman, but they took a long time to publicly explain why they did it. This obviously angered many people because on one hand, they tried to make people believe that their actions were right, but on the other hand, OpenAI itself claims that artificial intelligence is very dangerous and they will handle it responsibly. However, their behavior resembled an irresponsible and highly centralized organization.
Basically, it was a sudden announcement of this major decision and change without any warning or explanation. So in the case of the transition to Proof of Stake (PoS), it was something we have been discussing since early 2014. So people in Ethereum knew from the beginning that the transition to Proof of Stake was part of the roadmap. I think that was done well, to clearly indicate that it’s part of the roadmap.
Vitalik: I think one thing we didn’t do well is that we did underestimate the complexity of the transition. I think if we knew earlier that it would be this complex, we would definitely choose a simpler version of Proof of Stake to start with so that we could achieve the goal faster and shift the focus to really improving scalability. It was a mistake, and I think we have learned from it. And looking ahead, I release an Ethereum roadmap every year, basically trying to summarize the direction of Ethereum’s research and development team and what people are trying to do and implement.
Interestingly, the changes in the roadmap have actually become smaller. If you look at 2020 to 2021, 2021 to 2022, and then to 2023, and now this final change from late 2022 to late 2023, the changes are much smaller. Of course, there are still some changes, and I think a major change is that we have to find some solution to the centralization of stake, but other than that, there aren’t too many changes. So I think it is definitely an important thing to handle the community aspect of the remaining major transitions well.
Dao Qu: Then we move on to the fourth and final part, about the future of humanity, imagination, and the contours of the future. What impact do you think AI will have on human society, and what role can Plurality play in it?
Vitalik: Obviously, artificial intelligence itself has many risks, right? I mean, if you create a super intelligent artificial intelligence that is different from humans, it could be a huge risk to human survival. Even if that risk is managed, there are other risks, such as whether it would enable previously completely impossible centralization of power and authoritarianism. And here, this is not just a theoretical issue. We have even seen the effects of modern machine learning, facial recognition, and surveillance cameras in detecting and suppressing protesters in many countries.
Unfortunately, these technologies have been continuously used against opposition movements in Russia in the past five to ten years and are now increasingly being used elsewhere. So the question is, the technologies we have essentially allow a company with 35 employees to serve a billion people. So is it possible that the same technology would allow political elites to control a billion people with just 35 people once it is implemented and cannot be escaped?
This is indeed a huge risk in all directions. I think it’s important to think about these risks. I think maintaining a focus on and openness to safety is valuable for all these reasons. I think even when considering brain-machine interfaces, I am very optimistic about them.
Vitalik: I think it can be really amazing. You can feel like you can understand every language and even calculate 20-digit numbers in your mind. But on the other hand, we are actually talking about computers reading your thoughts. Computers are created by large crowds, including companies like Google, Facebook, Huawei, etc. We have given a lot of trust to these centralized crowds.
So it is valuable to try to keep these things as secure, valued, and open as possible. I think for plurality, what I see is that in many cases, being able to reveal consensus that already exists but people are not aware of is very valuable, right?
Being able to quickly find consensus on some issues, for example, where should we subsidize or put extra effort into certain technologies at a certain time? If there is no good way to achieve consensus, then historically what has happened is that these roles have been filled by elites. And the more roles are filled by elites, the more they can exploit it to gain and maintain power.
So being able to achieve this kind of consensus in a secure way and really provide us with information that we didn’t have before can potentially make decision-making better. I think it can be very valuable for various decisions, especially when the world is changing extremely rapidly.
In the last decade before the changes brought by artificial intelligence occur, whether it ultimately makes the world incredibly good or in the worst case scenario, it becomes worse, our perception of the world may be completely overturned every two to three years. This will be a real challenge that needs to be managed well, and while experiencing these changes, keeping the world we care about is something we must work together to achieve. We hope Plurality can help in this regard.
Dao Qu: Thank you, Dao Qu is very grateful for Vitalik’s sharing.