Vitalik: Radical Market, ZK, Privacy and more

Vitalik

Editor’s Note: This article is a translation of the EthTrader Communiity video interview organized by Kobe van Reppelen, interviewed by VitalikThis translation only selects some questions. The video address and the full interview are attached at the end of the article.

General problem

Q: What do you think about the entire cryptocurrency community? On my side, it seems that there will be a lot of controversy between different gods. They not only hope that the technology/coin they support will succeed, but also hope that other projects will fail. Do you have such an idea, that is, the future may only be a super good blockchain? Or is there still enough room left for different blockchain technologies in the future?

A: I do think that there is still enough room left for different projects. Of course, no matter how many cryptocurrencies we have now, 15,000 or more, certainly not all projects will succeed; I think most of them will be completely forgotten in the next few years. But there is still room for more than one project to survive, and there are many projects within a blockchain ecosystem.

One of the benefits of coexistence of multiple cryptocurrencies is our ability to discover and discern the differences and gains and losses between them. For example, like the different approaches taken by equity proofs, at the beginning we (Ethereum) emphasized the decentralization, emphasizing that nodes can be operated cheaply, emphasizing security margins and fines. But at the same time, there are many other types of equity proof systems that are built around the idea of ​​limiting the number of nodes; such systems tend to have faster processing speeds and block times, but on the other hand It may be easier to get into the center.

There are many similar gains and losses, such as centralization and efficiency, which are definitely one aspect, and functionality, performance, and simplicity may be another pair (the attributes that need to be weighed). Obviously, there are many kinds of consensus algorithms. Therefore, this is both a technical experiment and a social experiment, such as chain governance, non-changeable governance, and Vlad governance.

These things are very interesting, and the good thing is that there are so many different communities trying out so many different models. Of course, I also have my own ideas. I feel that some of these experiments will be easier to succeed. These views are also reflected in my design of the Ethereum agreement. But in general there is still room for debate.

Q: Both you and Glen Weyl believe that the cryptocurrency community is the Radical Market concept, such as the excellent test site for the second-party decline vote and the Haberg tax. Blockchain projects do require better rules, and different communities are basically more willing to experiment with radical reforms. If these ideas are really successful, do you think they are likely to be transplanted from the blockchain world to the real world? Have you foreseen that the incumbents in the existing system would reject these ideas for fear of losing power and property? How does the right idea overcome these obstacles?

A: I think the first thing to be clear is that exiting the blockchain world is not a prerequisite for joining the real world. It seems that I personally hope that some interesting Glen Weylian mechanism design ideas can take root, and that Ethereum and blockchain can finally bear fruit, so why not simply combine the two?

The Gitcoin CLR is one of the experiments I have long been interested in. Last month they did this round of CLR matching, where you can see a list of projects, and then everyone can donate money to these projects and determine the actual amount of the project based on Glen’s formula: donate individual participants Take the square root, add the square roots to find the square root sum, and then find the square of the sum of the square roots. This is the actual value obtained by the project; because this value must be greater than the total donated by the donor, then the difference in the middle is A center uses its own funds to make up.

I was very impressed with the results of this first experiment, because the results show that this mechanism works very well. Now Gitcoin is already in the second round.

When these experiments are expanded, the problem we need to counter or say is that these experiments will be attacked; one of the things I have been paying attention to is that this mechanism has no inherent conspiracy resistance. If someone can easily create 10,000 accounts or someone can easily bribe others, the whole mechanism will collapse. In these cases, if the collusion is very easy, most of these types of mechanisms will fail. So people need to think of ways to ensure that people can’t easily create fake accounts and conspiracy.

This may introduce many measures that people in our circle do not like, but this is the problem we need to deal with. One of the dark side of the cryptocurrency circle is that these completely anonymous attackers are usually very intelligent and resource-rich attackers, such as those who launched a denial-of-service attack on Ethereum more than two years ago, and those who are black. DAO and other contractors. I am pretty sure they will also start with the CLR. Given this situation, if we determine that a certain method can make the CLR stand up in this environment, if the society ultimately wants to use this mechanism in some large-scale environments, then they can use the same root as the CLR. A homologous scheme. Is this not a matter of course?

How long do you think we have an identity agreement with good anonymity? What is the most promising technology to solve this problem?

A: This problem is more difficult. To create a unique identity system is like creating an infinitely secure system. I am not even sure if this is possible.

Because, even if it is based on the government’s human identity system, what happens if we create a global standard and issue an electronic passport with digital signatures to everyone? Well, North Korea will print out 30 million passports, and then all passports will go to Kim’s drawer. He will use these 30 million passports to collude with others and then search for all the money in the CLR. Then North Korea has become a world power. Uh huh.

Therefore, it is very difficult to create a system that is really robust enough to be abused. For some people from traditional fields, it may be counter-intuitive. If the decentralization scheme is really useful, it will be more useful than the centralized solution, because the centralized identity scheme seems useful, but you know that this system may be destroyed. May be hacked. Recently, there have been many news reports that tens of millions of national identity information in some countries have been stolen by hackers (of course I have forgotten the specific figures). If a malicious government finds this to be a way of making money, they will undoubtedly establish a secret agency to operate this time.

I am really not sure if we can create a very, very strong identity system. But build a system with an increasingly high level of security, test whether it can work in a specific environment, rather than anti-counterfeiting identity functions – following this process will make us cheaper and easier to develop. . I will try to make it easier for people to develop an identity system: holding an identity through a filter is as natural as everyday activities, while making it harder for people to falsify.

There are many types of research that are going in this direction, and we may be able to see some systems that can operate effectively under certain conditions in the next few years. But frankly, I don’t know exactly.

Ethereum

Q: Do you think that the governance of Ethereum has failed? What are the main advantages of Ethereum’s unique governance structure?

A: I think that if you really see the wide-ranging wishes of everyone in this community, you will find that these results have been achieved.

For example, technical upgrades, from the front to the home, from home to Byzantium, to Constantinople, have been achieved one by one.

And when those who need some degree of centralization and rapid development, such as the DAO attack two years ago, people are united.

The hard forks can basically be seen in 3 to 6 days. The block rewards have changed from 5 to 3, and from 3 to 2 have succeeded. Serenity and Eth2.0 are also advancing, and Eth1.x is advancing. ProgPoW is slower to advance but is still going on, and people are also organizing audits.

From the results, at least as far as I know, the mainstream aspirations of the community have been met.

Perhaps only long-term performance is the best criterion for the quality of a governance system. Because if it doesn’t perform well, people will think it doesn’t make sense, as long as people who are not very idealistic will move to a better performing system.

I think we did a good job, although it is not perfect. I think the main challenge we face now is (and it seems to be where the signs of failure are to appear), and it seems that the participants in the governance process have spent too much time on sharding. But in fact, if you look at all the core developer’s meetings, ProgPoW has been discussing it for half an hour at each meeting (Translator’s Note: This paragraph is a bit messy, so you can be suspicious).

There is also the issue of capital recovery, which will jump out again in the same form every few months. These problems are repeated and a bit disgusting, and indeed cause a lot of burden on many people.

Part of the reason for this seemingly is that the way in which the community’s opinions are gathered is not trusted, or that there is no consensus in the community that there is legitimacy. For example, now that we have Carbon Vote, you can vote on GitHub, but you can also vote on Reddit and EthTrader.

I personally value these signals very much, especially when multiple signals pass the same information. For example, what I found interesting is that the recent collection of opinions on financing with inflation is still in the early stages of the collection, with only 90 people involved, but many support this proposal. I haven’t seen a very strong negative opinion, at least not receiving the kind of strong opposition that EEP 867 (the universal fund recovery program) experienced. But the interesting point is, I guess you never thought that those who opposed the money recovery may also support inflation financing! But, well, this proves that people might think like this.

I want to add tools that better reflect the views of the community, not just a small number of rich people who can vote, or everyone like Reddit can come in and vote. (If you have such a tool) that would really be useful. I think EthTrader’s polls are also very interesting because they use donu to govern, which means you have to participate in the community to get voting rights. This model can make a big difference. There are many experiments like this.

One of the experiments I liked most recently is that I think people should independently list the GitHub accounts or Ethereum addresses that they think are active members of the Ethereum community (and reach a certain level of security), or something else. Corresponding account. Jeff Goldman recently made this suggestion, I pushed him back, and he also felt that my pushback was pretty good.

The idea behind this kind of experiment is that once you have the list of people, you can use these lists as input sources with a higher level of security, in the CLR or in community polls, declining voting. , used in all such different forms of voting. It is an improvement compared to the Carbon Vote of the favored wife and the Reddit vote that is easily kidnapped.

But such a plan is not without problems. Because judging who is part of the community is itself a terrible subjective practice, and it is highly likely to be affected by prejudice. It may not work in many cases. But I basically think that there are more indicators than the lack of reference.

Q: How do you protect your identity in Ethereum? What is the current progress of privacy protection technology? When can I see more friendly privacy protection technology?

A: From a technical point of view, it is easy to establish a privacy-protected system on Ethereum. But as far as the facts I have observed so far, the main obstacle to protecting privacy in these systems is that in Ethereum, you need to use Ether to pay transaction fees. For example, you include a zero-knowledge proof in a transaction that proves that you own the ownership of Ether that is invested in the privacy protection system. But the deal itself links your trading output to the Ether you already have, and it unlocks your anonymity itself.

This of course is where the account model needs improvement. In theory, we can fix this flaw through a Layer-2 network, but no one has yet done this. I have been encouraging people to develop such a network. In addition to the privacy of money transfer, another reason I support people for this development is that there are many non-financial applications in Ethereum. You certainly don’t want some of your activities to be associated with all of his other activities.

I think the best default setting (of course, just the default) is to automatically assign a different account for each difference. But if you really have such a design, you need some way to migrate Ether between different accounts, and you can’t let others see that you are right-handed on your left hand (and then associate all your accounts).

So all in all, the privacy protection solution has a very good initial application scenario.

Q: I think Ethereum’s initial vision/plan included EF to fund the development of a Trinity technology (blockchain, Swarm, Whisper) and a browser that uses these components. So, is this plan changed now? If it does change, do you think this is good or bad?

A: I think this plan (making EF as the fund sponsor of these three technologies) has been abandoned to some extent, but the result is good, because the result is not that the three technologies themselves are abandoned, but developed. The work was transferred to other community members. For example, some people have integrated Ethereum into the Opera browser, as well as MetaMask, Brave browser, HTC blockchain mobile phone, Samsung blockchain mobile phone, and so on. There are already many ways for you to use Ethereum, and such a path is still emerging.

So I think the result is still good. The browser that provides users with an interactive interface has already begun to develop independently; Whisper’s progress is slower than expected, I hope that some things can be done as soon as possible; as for Swarm, it seems that the multi-file storage protocol is progressing well, so I am very happy.

Ethereum technical issues

Q: You have been thinking about randomness in the chain for a long time, from Randao in 2016 to VDF and threshold signatures today. So, in addition to the protocol layer security, what do you think can be changed by a cheap and usable arbitrarily random solution? With the randomness of the chain, what is the difference in the future world?

A: In addition to improving protocol security, I believe that easy-to-obtain random numbers are very useful for many applications. The most obvious is the lottery on the chain. You think about it, the millions of dollars of national lottery tickets, the source of entropy is opaque, and people can hardly trust them. If we have a robust VDF ecosystem with properly updated VDF hardware, the lottery is something that can run on the blockchain, and it can even be said that it should be done on the blockchain. In fact, lotteries and any type of game that contains randomness can do so.

There are other protocols that are quite random. There are some games that mix randomness and skill; there are many different types of mechanisms that ultimately require randomness because they don’t have full certainty. So the long tail of the random number is very large. You can even use it in the equity proof algorithm of Ethereum Layer-2.

There is still one! Improve the efficiency of STARK. Stark is now 50 to 150 kb in size. The reason for this is that Stark needs to use a protocol to create a Merkel tree for the extended computational process. Then the Merkel group itself is used as a random number source, let us randomly take out some branches of the tree and verify the correctness of these branches. Use this protocol for counter-cheating: prevent most of the proofs you create from being wrong, and only a small number of your carefully selected branches are right; to cheat you have to constantly try to create proofs until you find them. A proof of meeting the requirements. The advantage of using VDF is that after using VDF as a random number source, you don’t even have to try the cheat space. A Stark will be reduced from 80 branches to 20 branches, for example, so that the data size will be reduced by 4 times; so Stark for any operation can be easily stuffed into a block.

So there are a lot of random application scenarios, and there are many scenarios to unlock.

Q: Ideally, which three things can help Serenity deploy faster in the second phase?

A: Is the promotion of the second phase of development? I think it is mainly a long research and development work.

If you are technical and interested in participating in testing, verifying detailed documentation, adding tests, checking the legitimacy of tests, contributing code to a client, etc., these tasks can accelerate the development of Phase 2.

If you are a dApp developer, I would love to hear your opinion on the Phase 2 design. We are collecting your opinions, such as on the smart contract execution layer, what cross-chip transactions should look like, what level of account abstraction should be abstracted, and how the ERC-20 token standard should be designed in Eth 2.0. I very much hope that everyone can share their thoughts.

If you want to donate money, that’s fine. Now GitCoin is conducting a second round of CLR experiments, and a small amount of money you donate will be magnified 5 times and 10 times into project funds. You can go to support Lighthouse, Pegasus, ChainSafe and other great 2.0 teams.

If you are just a community member, you can also educate and support developers. There are still many things to do.

Similarly, if you want to ask if there is any way to speed up the development of Layer-2, because the Layer-2 solution like Plasma and ZK-Rollup is also very important, then please get started and give your feedback. Because those tests are ready. Another thing is to help standardize and ensure that we don’t develop 15 incompatible protocols and then drive the availability of these projects.

(Finish)


original material