Scroll Talk #1: Scroll goes hand in hand with ZK

Source: Scroll CN

Scroll Talk is a podcast series hosted by Scroll CN. We will talk to the Scroll team and the Scroll ecological project through different forms to help everyone understand Scroll better.

In this episode, we asked Ye Zhang, co-founder of Scroll, to talk with him about Scroll and ZK, including the design and trade-offs of zkEVM, the choice of proof system, hardware-accelerated prover networks, and the future of ZK.

**F.F: Hello everyone, welcome to Scroll Talk, today I am very pleased to have Scroll co-founder Zhang Ye, we at Scroll CN have also published many interviews and speeches about Zhang Ye. Then this should be the first face-to-face meeting with Ye, and then first thank you very much for Ye coming. In the entire zero-knowledge proof community, Ye is now very influential, but we still want Ye to make a simple self-introduction first. **

Ye: Hello, thank you Scroll CN for arranging this interview, and then I have always thanked Scroll CN for his contributions in the Chinese community, including the quality of the translations, and helping us have a great influence in the Chinese community. A simple self-introduction, hello everyone, my name is Zhang Ye, and then one of the co-founders of Scroll, my main direction is zero-knowledge proof related research. I previously focused on three directions.

The first direction is the hardware acceleration of zero-knowledge proofs***, which began to be done about 5 years ago, because one of the biggest bottlenecks in the use of zero-knowledge proofs 5 years ago was that the generation of proofs was very slow. For example, it takes an app like Zcash to generate proof for a transaction, perhaps 10 minutes or more. This then leads to the argument that zero-knowledge proofs cannot be adopted by many systems because the proofs are too inefficient. So my first research direction was how to accelerate proof generation with GPU, FPGA and ASIC hardware.

The second direction is these cryptography and mathematics **** in the back end of partial knowledge proof. Because zero-knowledge proofs are a very complex cryptographic protocol that involves a lot of mathematics such as polynomials. Then my main research work is to read a lot of papers to see how to optimize some of the existing algorithms, which will be more theoretical.

The third direction is more application level ***, that is, how to design the architecture for zkEVM, design the circuit, how to generate proof for zkEVM.

In general, it covers three directions, hardware acceleration of zero-knowledge proofs, theoretical algorithms of zero-knowledge proofs, and related applications of zero-knowledge proofs. ***

Then at Scroll, I will mainly focus on some research-oriented work, including zero-knowledge proof-related research, protocol design research, and then some company-wide related strategies.

**F.F: Thank you, Ye, and then we know that you have been doing ZK research, what kind of motivation made you choose to start Scroll, and then have been deep in the ZK field, what motivates you to keep going?

Ye: It’s a very different story. Now most people have heard of zero-knowledge proofs or learned ZK because they realize that blockchain has the need for zero-knowledge proofs, but in fact, my learning process is the opposite. I actually did the opposite, I was first attracted to ZK, and then found that ZK can be used in the blockchain. When I was an undergraduate, I was doing some research related to hardware acceleration algorithms with a senior in the lab. Then the hottest thing at that time was actually doing AI acceleration, but I was not very interested in AI, I felt that his parameter adjustment process did not have a mathematical model that I could understand, that is, why this parameter would be the result through training. I prefer more deterministic mathematics, where I can know the probability of something happening. So I would naturally prefer the cryptography number theory, and then I discovered the algorithm of zero-knowledge proofs at that time, and then found that it had a very large hardware acceleration requirement. So I started to do related research, and later, in the process of studying acceleration algorithms, I found that the algorithm itself will be more attractive than hardware acceleration, because it involves a lot of very clever polynomial construction, some protocol construction, and so on. If you really look at any zero-knowledge proof protocol, you’ll see that it’s really ingenious, through the design of polynomials, to encode programs. Then pass some polynomial points to verify some properties of the polynomial, and finally compress it into a very, very small proof, the whole mathematical construction is very ingenious. So I initially entered the zero-knowledge proof industry completely attracted by the charm of his mathematical construction, and then later found that the thing I was studying just happened to solve the biggest problem that blockchain has now encountered in essence, which is scaling. ***

Then I later realized that the entire Ethereum ecosystem is very prosperous, and it has a very good open source community, which is more in line with my own personal beliefs, and its entire research atmosphere, embrace open source attitude, and pursue academic rigor style also made me completely attracted to it, and at the same time realized that the whole blockchain is not a hanging story, but a real architecture that can solve problems in many people’s lives. It could be the next generation of financial infrastructure, and a lot of people really need transparency, they need censorship resistance. So I think blockchain has real application scenarios, and then at the same time my technology can solve this problem.

At the beginning of 21 is the best time node, because the efficiency of zero-knowledge proofs has increased by 2-3 orders of magnitude. When a technology has 2-3 orders of magnitude improvement, there will actually be huge opportunities, whether it is entrepreneurial opportunities, or other new opportunities. Because what couldn’t be solved before can now be solved. At the time, I felt that zkEVM was actually the biggest opportunity, and there were not many people doing it, or not many people doing it. Then we had a very good opportunity, coupled with the accumulation of such technology, to solve such a problem, so we started to do Scroll.

In fact, before I studied for a doctorate, I was also doing zero-knowledge proofs. But I realized that one of the things I realized was that you are in the industry, for example, at Scroll, and you have a lot of flexibility to do ZK related research, whereas in school you have to work with a mentor, and then you may only be able to work in one direction.

But with Scroll, you may have more flexibility, because the first problem you solve is a real industry problem, so after you solve the problem, its impact will be greater. And then second, you won’t just be limited to one area of your Ph.D. research, you can collaborate with more people through some grants, through some other ways. So words. In fact, at Scroll, I am doing the same thing, but in the industry it will be more influential, solve the most real problems, and the scope of cooperation will be wider. So I think this path will be more attractive to me than doing a PhD.

ZK Technology Development and Future

**F.F: Thank you Ye, who understands the fascinating charm of the math behind ZK, and then attracts you to do related research all the time. One of the main breakthrough points I heard should be two or three years ago, is ZK similar to this year like ChatGPT came out, there are a very large number of breakthroughs. **

Ye: Yes, yes, I think so, but it’s not like ChatGPT is a flashpoint and then suddenly detonates everything. It is a process of superposition of many layers of factors, such as the hardware acceleration direction I have been working on before, which can increase the efficiency of zero-knowledge proofs by 10 to 100 times; Then add some new representations of the circuit by polynomials, such as more efficient representations through higher-order custom gates and lookup tables, and reduce the overhead by a factor of 10; Then there are some recursive proofs that aggregate many proofs and save a lot of costs in validating that piece. So I think these three points add up to a huge efficiency gain.

In the end, of course, the result is an efficiency gain, but it’s not like ChatGPT is a sudden explosion, but the result of the efforts of cryptography experts and the efforts of many hardware engineers. ***

**F.F: Since we talked about AI, what do you think about the combination of ZK and AI, including that Worldcoin has been released, using ZKML technology, what do you think will be the difference between ZK, AI and the two respective fields of ZKML? **

**Ye:**I think *** There are indeed many people doing ZKML now, but I think this direction is still quite early ****, it has some application scenarios, such as identifying whether the photo is really taken through your camera, there is no complicated PS; It can prove that the audio is not a person; It can be shown that Microsoft is not providing the same model to everyone, because when you give the model an input and it returns you an output, you can’t be sure if it’s a different person changing the model. There will be such small application scenarios, but I have not seen a particularly large demand, why ZKML must be such a widespread application as AI. ***

Because for example, like ChatGPT, most companies that own the model are absolutely powerful, and it doesn’t have to prove to you that I have to use the model, and you can’t ask him to do it. I think unless it’s happening in a market where there are 10 companies like ChatGPT, and ChatGPT doesn’t want to do it, then one of the other companies is willing to do it. A wave of users with such demand will choose the services provided by this company. But now it’s a market of just a handful of companies that can do something like ChatGPT, and then they don’t have a strong incentive or incentive to do that for you. So I think this road is still quite long, plus a lot of things like photos or audio, there are still many problems that have not been solved, and you may also need some hardware to build this system.

Overall, I think there is still a long way to go, of course, I think ZKML will have some new gameplay strategies in liquidity management, and then there may be some small use cases, and larger application scenarios still need time to test to find a product that matches the market. And ZKML it actually can’t prove that the training process is right, it can only prove that an impact factor is right, so this further limits what it can do, I think there is still a certain way to go. ***

Most of ZKML’s companies are still developing tools. I know that some ZKML companies are figuring out how to convert code written in TensorFlow or Pytorch directly into ZK circuits and directly generate proofs, which may be an interesting direction. Starting with DSL, SDK, and then encouraging everyone to have new innovations, are still very early. I think it may eventually develop into ZK like general computing, but some algorithm libraries more suitable for ML, such as some matrix multiplication or convolution, will be more conducive to such applications, rather than just ZKML, I think there will still be a long way to go.

Then there’s Daniel Kang, a professor at UIUC, who we’ve invited to give a talk at Scroll’s ZK Symposium, so if you’re interested in this direction, you can check out our series.

**F.F: Okay, thank you Ye. Then I feel that ZKML, or relatively early, is now some relatively small construction directions, involving general computing is still relatively early, may get the later market is very mature, when we have privacy needs, ZKML will have its place. Looking at the broader scope of ZK, Vitalik made a statement before that roughly means that ZK and blockchain are equally important concepts, what do you think of this view? **

Ye: I think it’s true, because ZK really solves a lot of problems that blockchain can’t solve, and it’s a very perfect combination. Blockchain cannot solve the problem of scaling, ZK can compress the calculation and solve the problem of scaling. Blockchain is always open and transparent, it can not solve the problem of privacy, ZK can hide information, can solve the problem of privacy. So I think ZK and blockchain are a naturally very good combination. ***

In addition, ZK’s support for general computing is also very rapid, so I think it has a very big opportunity. For example, from the privacy aspect, like private transactions, privacy pools, including some on-chain poker, you don’t want to deal cards on the chain, others see your cards, you can hide information through zero-knowledge proofs. Games like this that hide information, you can only do on the blockchain through ZK. In terms of privacy, ZK Identity is also a very interesting small direction, very promising, how to let a billion users use the blockchain, then we may need to make some existing identity systems ZK, and then everyone is willing to put some information on it.

Then scalability-related words, such as various rollups to compress calculations, various coprocessors to compress some calculations, and finally put the proof on the chain, so that it is a very good combination of on-chain and off-chain.

There are also some very interesting small directions that are promising, and there are teams that are building some ZK cross-chain bridges, or ZK hardware to provide services. But I think it’s still a few years away from being particularly mature. Whether it is from the convenience of the developer SDK or the efficiency of ZK, there is still a long way to go in terms of security.

**F.F: Understand, thank you Ye, listen to your description, ZK and blockchain are equivalent to a complementary relationship, so in addition to these application scenarios just mentioned, from the perspective of efficiency and fairness, what changes do you think this technological innovation will bring to the real world? **

Ye:** I think it’s a very strong feature to be able to make any computation trustless. You can throw any computation on a platform that you don’t trust, and then have it return you a result, generate a proof, and prove that its result is correct. This ensures that your calculations are correct and verifiable. And then other applications like I said are identity, privacy, scalability and so on.

**F.F: Understand, thank you Ye, that is, ZK it may have a blessing on general computing, whether it is to provide privacy or credibility, is a very promising direction, if you do not do Scroll, let you start a second business in the ZK field, which track and direction will you choose? **

Ye: This is a very difficult question, I think *** first of all, zkEVM is definitely the largest direction, because it carries the Ethereum traffic entrance, it must be the largest ***. If I have to choose other directions, I am personally quite optimistic about making coprocessors, and can do non-EVM calculations very efficiently and verify. On the other hand, I think it is very difficult to build a good identity system, and it can solve many real-life problems, especially when I went to Africa before, I saw a lot of problems caused by the immaturity of the financial infrastructure, so I think identity will be a big direction.

Then if I personally must choose, I think *** If the volume is not very large, I think the opportunity of identity is great, and then if it is a very strong engineering team, it must do more complex things, I think this ZK coprocessor will be a better direction ***, but now there are too many people in this track. So I think identity will be a track that has not yet been popularized, and I think it is not only a direction that requires technology, but also business strategy. You have to think about which business partners you want to work with, can you directly ZK their large amount of data, so that you can expand your user base more quickly, technology may be a smaller problem.

If you are a very innovative person, you can also try the direction of ZK games, the game needs you to have a good design, ZK the information that needs to be hidden. But ZK is not a one-size-fits-all tool, and it cannot solve all privacy problems. It will require the prover to know certain information. So I think this game needs to be designed very cleverly in order to use ZK to the extreme. If you are a person who has a lot of ideas and likes to play games, thinking clearly about your game logic and creating a fun ZK game is also a very interesting direction.

F.F: Thank you, Ye. You just mentioned three directions, the first is the coprocessor, similar to the direction that Axiom is doing; The second is the direction of identity, which can be understood what Worldcoin is doing, is one of the special examples; The third should be the game, which will be a daily contact direction for ordinary users. Then Ye you just mentioned that you just returned from Africa, so by the way, we want to ask you this time to go to Africa to promote the technology of ZK including Ethereum, what are the gains?

Ye: This was a very unique experience. To give a brief background, in February this year, Vitalik and the Ethereum Foundation’s Aya and others traveled to 4 countries in Africa. They probably spend close to a month holding events with African communities, and then meeting some of Africa’s founders to learn about the continent. Because Ethereum is actually a relatively small community in Africa, they go there to learn about some of the current state of the community. What does the community need now to spread this value of Ethereum. They came to the conclusion that Ethereum was still a bit expensive.

They hope to arrange a trip to Layer 2 that will bring Ethereum’s Layer 2 to Africa. Because people in Africa, they can’t afford Ethereum, so they can only enter the Ethereum world through Layer 2. So in April and May this year, I also met Joseph, the organizer of their trip, through Vitalik, to see if there was a chance to say that I could organize a Layer 2 trip. After we talked, we felt that our values were very aligned. Because part of Scroll’s values is that we want to bring real users and use cases to the blockchain, we’re excited to learn about some of the real needs in Africa.

After going there, I will find that it is really different, which makes me more confident in the real use cases of *** developing countries and emerging countries. Because in fact, before going to Africa, many people, including those I am in contact with now, are still questioning, that is, whether blockchain is a real demand. It’s not just a scam, it’s just a tool for everyone to issue tokens. I think people who can say such a statement are actually because whether in China, other parts of Asia, or in the West, the United States and Europe, everyone’s understanding of blockchain is giant whales, liquidity mining. It’s not that they really need blockchain in life, but they just feel that saying that there are some tools on it that can make him more profitable, and sometimes his property may be safer on it. It is not a particularly needed tool.

The two countries we went to were Kenya and Nigeria, and it was clear that everyone there really needed blockchain as a platform for their daily lives. A very obvious example is that there is no way to transfer money directly through a bank to make a transfer between two neighboring countries in Africa, and it takes a large circle to reach the account. Because their financial infrastructure is really poor, and they are completely unable to build a global support system.

So they really need a payment tool in the first place, and blockchain is very useful just as a payment tool and can really change their lives. Because they need blockchain payment medium if they go to other neighboring countries. Many people say that what blockchain can do, just a global payment system, sounds like a very single use. But the global payment system can solve the needs of many people, especially in countries where the financial infrastructure is not so perfect. But because you may be in China, the United States, Europe, where the infrastructure is very good, you don’t usually worry about such problems.

The second thing is that their inflation is very high. Their currency may have been 10% inflationary since we got there. Imagine that the RMB or US dollar in your hand will depreciate by 10% after a month, and financial management may only rise by 3~4% a year, and then prices have been rising. It affects their lives, and then stablecoins are a way for them to get dollars, they need dollars, because dollar inflation is low, so they want to get dollars, but they can’t get dollars, because they can’t open an account with a U.S. bank. So they actually get USDT by buying dollar stablecoins and then holding some on-chain assets, which is a very important way to prevent hyperinflation. Maybe everyone can hold RMB in China, and USDT is only needed when buying cryptocurrencies. But they are needed in real life, and they frequently transact OTC and convert them into their own currency when they actually use it. So I think it’s a big application scenario, and in these countries and a lot of other places, they really have that need.

And then the third thing, because of the imperfection of their financial infrastructure, led to their credit rating and identity being very imperfect when borrowing. So they borrow money, let’s say $100, it can take a month, and they need various approvals, because information does not flow between various financial institutions. So this leads to the fact that *** lending, as a very large business of banks or many financial institutions, is very imperfect in their place. So I think it’s also a huge opportunity. ***

In Africa, there are many real application scenarios that require blockchain. For example, if there is a good identity system to solve their problems, can provide them with some loans or other services on the chain, I think it will be a very valuable thing. This is the first time I feel that your technology is really changing the lives of people in many corners of the world, and that’s a very important thing.

Part of Scroll’s values is that we want to get the next billion people into Ethereum. People will often complain that BSC is very non-decentralized, Ethereum is very decentralized, but ether is expensive. And then there are a lot of real users on BSC who are just playing, because there’s Binance, and then Africa is what I saw for the first time, there are a lot of people who are really using Binance to make payments, because it’s very simple and easy to use. We hope to bring this part of the real users back to Ethereum, this is part of our mission, we want to bring the next billion users back to the more trustless Ethereum after reducing fees through Layer 2. Because if you keep your money in a centralized exchange, there may be some problems. So we hope to be able to put it on a Layer 2 and then inherit the security of Ethereum, which is a great opportunity.