Exclusive interview with Elon, COO of Epik Protocol: Using blockchain technology to build a decentralized, co-constructed, shared and mutually beneficial secure and reliable knowledge graph collaboration platform

On May 21, the Lijiang Crazy Water Festival, the first year of computing power nationwide tour, was successfully held, hosted by Wabi.com, Caiyun Computing Power, and Interstellar Vision, with the chief cooperation of Bitmain and the strategic support of BitDeer. During the meeting, Kiki, co-founder of Wabi.com, conducted an exclusive interview with Elon, COO of Epik Mingshi Protocol. The following is the transcript of the interview:

Kiki

I would like to ask Elon to give us a brief introduction to what the EpiK Mingshi Protocol is and what problems it aims to solve?

Elon

EpiK Mingshi Protocol is a collaborative platform for knowledge graphs, a decentralized collaborative platform for knowledge graphs. In other words, we are doing blockchain + AI. The 14th Five-Year Plan stated that the next generation of AI must have learning and basic judgment capabilities. We will think: What does this AI need to do in order to move from the current AI to the next generation of AI? It depends on our knowledge graph. Of course, knowledge graphs are also widely used in life. Every time you buy something on Taobao, smart search and smart recommendation are a knowledge graph; including Siri, Xiao Ai, etc., which we are using, they actually have knowledge graph architecture at the bottom. The EpiK Mingshi Protocol project is to convert the content of various industries into this kind of knowledge graph and store it in a distributed manner. So we also belong to the mining sector, because the knowledge graph involves the concept of distributed storage. After we produce the knowledge graph, we deliver it to our miners, who help store these knowledge graphs, that is, miners need to bring servers to participate, and the server is equivalent to a mining machine.

What problem does it solve?

The EpiK protocol will develop towards the AI ​​intelligent era in the future. It is known as the fourth industrial revolution. The entire society will become more and more intelligent, and the AI ​​part will become more and more important. Of course, it will become more and more dependent on knowledge graphs. Therefore, our EpiK protocol is now equivalent to making an AI database.

Kiki : What are the roles of the participants in the ecological mechanism of the Epik Protocol, and how do they work together?

Elon : There are several main participants in our ecosystem. The first is the domain expert, and then the bounty hunter. These two are the people who produce the knowledge graph, and then hand it over to our miners to store it in a distributed manner. The real user is the knowledge gateway. In the vast AI companies, the knowledge will be called through the knowledge gateway platform. Each role in the entire ecosystem has a pledge mechanism to ensure the smooth development of our entire economic model. What the domain experts have to do is to transform the content of various industries into the architecture of the knowledge graph, so that AI can understand it, it will have the ability to infer, and it will become more intelligent. Let's take an example of what the domain experts are responsible for: turning the book Romance of the Three Kingdoms, a novel of more than 10,000 words, into a knowledge graph architecture, how to do it? The domain experts of Romance of the Three Kingdoms have to list the framework of this book. After listing, they will find AI data annotation companies through normal channels, and use a large number of trained AI labelers owned by AI data annotation companies to do artificial intelligence annotation, but the whole process is repetitive and single. Now we are using the currency method to do it. Domain experts only need to build the framework, that is, to split each chapter and paragraph of the book Romance of the Three Kingdoms into simple multiple-choice questions or fill-in-the-blank questions, just like asking you whether Zhang Fei and Zhang Yide are the same person. Some bounty hunters who want to get the money will go online to check. If they are indeed the same person, then I will label them as the same person. This is called disambiguation in AI. They have to do a lot of such things before they can build the final architecture of the knowledge graph. After it is built, it will be given to miners. Then the two ecosystems I just mentioned need to be pledged. The pledge of miners may be the heaviest. That is, each miner needs to pledge one thousand to two thousand tokens to us to get a good computing power. The whole economic model depends on this. Finally, I would like to mention that the miners pledge in the currency that can be locked in. However, the coins we mine every day are released slowly in 7+7. The coins you mine on the first day will be fully released to you in seven days on the eighth day. This is also conducive to the increase of our overall tokens. Now we only need machines with 8, 16G, 256G and SSD. This configuration is very low for mining. Any slightly better computer at home can come in to mine. Then, with this as an anchor point, a large number of miners will come in. They need to pledge, and to pledge, they need to buy coins on the exchange, which will increase the value of the token.

Kiki

What is the framework and design concept of the EpiK protocol’s token economic model system?

Elon

In fact, it is very simple. Every participant just needs to do what is best for his own interests. Because what we have designed is that everyone does what is best for their own interests, but at the same time they can work together to bring the entire project forward.

Kiki

Mr. Elon, what kind of product is the Knowledge Continent launched by Mingshi Protocol, and what is its relationship with Epik Protocol?

Elon

I actually mentioned just now that we need to produce a knowledge graph, and then we need to split the book Romance of the Three Kingdoms into small questions. Where should we put them? Then we will produce a knowledge continent. You can see it. He will put the questions here, and he uses a gamification method to deal with it. First, you play this just like playing a game. You have to kill monsters and upgrade. You complete the questions every day, and then you will have points, which can be converted into RMB in the end. After the main network is launched, we have predicted that the hourly wage for each person playing this will not be less than 30 yuan, so this will also attract many people to play. Some people are just like playing a game, just doing questions, and their food money for the day is back. We will put the questions in the knowledge continent, and on the other end, anyone interested can choose various types of questions to complete, and all will be collected on it. We are also talking to some game companies now, because some small game companies are in a very difficult situation. First, they cannot make any big games. Second, their only form of income is from selling advertisements. Ads will pop up when you play the game, but this greatly affects the experience, so all small games are not doing well.

So we are considering cooperating with small games. They put the data we want to collect into their games. Maybe when you want to get a certain item, you have to answer first whether Zhang Fei and Zhang Yide are the same person? Then you can get the item after answering. What does the game company get? The token we give it, so it can survive, and the whole ecosystem can also participate together.

The first version we are making is a cloud library. For example, this one is about Tang and Song poems. This is a question answered by a customer. I will help him to calibrate it, which is equivalent to checking whether he reads it correctly. For every question you answer, there will be 10 other customers to calibrate it for you. In this way, the final result will be better. We adopted the needs of a professional Tang and Song poetry AI company to make a Tang and Song poetry reading companion software for primary school students. Then some voice data is needed to train its AI, so we help him do it, so our feasibility is relatively strong now.

Kiki

What changes will the Epik protocol bring to our lives and work after it is completed?

Elon

This brings us to the question of why we should use blockchain to do this, because the knowledge graph has a big pain point in the industry, which is data monopoly. I just mentioned a small case in my speech, which is equivalent to saying that to build this kind of knowledge graph now, you have to hire a large team, and the salary of each person is not cheap, so only large companies have the ability to support a team to train its knowledge graph, so as to train its AI to become smarter. So we say that in the future, in life, a lot of AI may be in the hands of large companies. But large companies are centralized institutions, and they are destined not to be too open and transparent.

Our future lives will definitely become more and more AI-based and intelligent. AI is equivalent to playing the role of a personal assistant. At the lightest, you may ask it which Sichuan restaurant is better for you to eat today. More importantly, you may ask which insurance is really suitable for your child. If you are sick, you may ask AI which hospital, department, and doctor have a better reputation. If we use Baidu AI and Baidu is doing bidding ranking, then the hospitals pushed to you at this time are those in the Putian system. Therefore, when our future lives have to rely heavily on third-party databases, we actually have a considerable demand that the third-party databases are open, transparent, and cannot be tampered with. This is why we use blockchain to do this. Therefore, the impact on our future lives is equivalent to the fact that there will be quite a lot of AI processing in our future lives. If the knowledge base behind it, that is, the knowledge base that trains its AI, is an open, transparent, and cannot be tampered with blockchain, then we will use it more assured and really have the data in our hands.

On May 21, the Lijiang Water Festival came to a successful conclusion. Thanks again to all our partners and media friends for their strong support.