Sep 14, 2018
PlatON’s founder Lilin Sun gave a speech on The 4th Global Blockchain Summit
On September 11th, the 4th Global Blockchain Summit.sponsored by Wanxiang Blockchain Labs was launched at W Hotel on Shanghai Bund. Dozens of important guests from home and abroad attended the lecture and attracted more than 1,500 visitors.
Lilin Sun, the founder of PlatON, was invited to attend the summit and delivered a wonderful speech on “Trustless computing-Next Generation Computing Architecture”. He shared the exploration and achievements of PlatON in the direction of high-performance Trustless computing network on the summit.
The following content is organized from the speech of Sun:
Before the speech begins, let me briefly introduce what PlatON is? In July of this year, we released the PlatON program in San Francisco. PlatON outlined a high-performance Trustless Computing network.
The blockchain is often considered the connector of the next generation of value Internet, connecting all industries. But in fact, with the development of the third generations of the Internet, we have encountered many problems.
First, the traditional Internet architecture faces a immediate challenges: the number of nodes participating in global Internet computing is growing, and exponential growth is beyond our imagination. Each node has its own computing power and storage capacity, and it will hope to get more computing rights and get more computing value, and the traditional architecture is more and more difficult to support this demand.
In addition, each node is generating a lot of data, but the data does not really flow. Traditional centralization organizations or corporate entities increasingly believe that data is their core asset, and even if they are paid, they are not willing to transfer data, especially private data, which leads to the inability to fully utilize the data. And in today’s market, no one can reasonably price data. How much is the data value and how should it be sold? The current Internet infrastructure cannot provide a solution.
In the past three generations of Internet, Web 1.0 generated the C2C business model. Web 2.0 made simple payment. Today, new computing architectures and technologies are needed to solve the problem of data mobility. Let everyone ensure that their data privacy is granteened, and that data can still flow freely without being taken away.
The blockchain provides a good idea for solving these problems, but it also faces two major challenges: scalability and privacy. The high-efficiencyTrustless Computing network that PlatON is building is trying to effectively address these two challenges.
In terms of privacy, we have been researching and spreading the MPC algorithm. In the past two years, we have built a general complete two-party computing architecture, including new compilers, virtual machines and a complete set of algorithm implementations. At the end of this year or early next year, a three-way computing common-purpose computing architecture will be introduced and used throughout the market. It is a very interesting logic, it may be more technical, the computing architecture has reached the circuit level, and it is executed inside the circuit.
The value that MPC can achieve is that, under the premise that local data is not collected and privacy is not leaked, the parties can still obtain a common result by performing the operation of the given logic. Of course, the technical challenges here are very large, the challenges of the compiler, the challenges of the algorithm, the challenges of the virtual machine, etc., all not easy to implement. Our team has continued to develop for two years and made such progress and will continue to carry out technical research in the future.
In terms of scalability, PlatON mainly uses VC algorithm. Now the blockchain system, every task, every node of each contract must run once, which is an important reason for inefficiency. In the VC architecture, the task computation is done under the chain, and only the computation result is chained, and the node only needs to verify the result. The essence of VC is to generate a very short proof, so the verification time required is also very short, which greatly improves the system efficiency.
Regarding the workflow of the VC architecture, I just mentioned a key issue. In the future, there will be countless nodes involved in the computation. In a Trustless Computing environment, I don’t know it and don’t believe it. There is no trust between us, but I need to give a task to it. How do I know if it really counts this task? Using VC to calculate the model, I input the parameter and function , it will return a result to me, and everyone can verify the result to validate whether it did a effective computation. This is fundamentally different from today’s cloud computing, consensus-based blockchain architecture.
Some people also asked us what is the relationship between VC and zero-knowledge proof? They have some overlap, but the focus is not the same. The VC’s core idea is to generate a short proof, which has been strengthened in the verification time, so it is mainly to solve the problem in the verification environment.
Of course, there are many verification solutions available in the market today. And the work we do is to choose the solution with the highest reusability and the highest global recognition among these kinds of solutions, as our support structure. We believe that this thing is like the infrastructure construction of telecom operators in the 1990s. Users don’t need to pay attention to what technology is used at the bottom, and we don’t have to stick to specific algorithms. How to use it, this is an engineering problem. Not a theoretical issue.
PlatON consists of four main parts: verifiable computing, privacy computing, scalable computing, and specific computing hardware. In the entire computation process of PlatON, parallel computing completely adopts the circuit mode, and all logic is written as a circuit. In the future, we will cooperate with mainstream hardware manufacturers in the industry to further improve computing performance in hardware.
This is our technical roadmap. This year we will launch a basic test network, and the main network will be online next year. At that time, the complete VC and MPC will be truly commercialized.
PlatON will have its own master chain released, but we are more of an infrastructure. So we welcome more links to join our network and call all our underlying services.
Many people are more concerned about our business model, especially what MPC can do, and here is a brief introduction. In fact, this thing is very interesting. Many friends will say that MPC is a magic, because I don’t understand how this happened, but it really happened.
All industries that have a large amount of data and value privacy protection can be used by MPC.
For example, brokers are often reluctant to give their data to each other when counting whitelists and blacklists. Because each organization spends a lot of risk control costs to get these data, it is impossible to contribute to the other party without compensation. If you pay for the other party, how to price? It is also difficult to determine.
At this time, the value of MPC is reflected. In the case where the respective data is not collected, everyone calculates a result together to complete the user’s portrait. In addition, it can be widely used in medical, joint credit, global scientific computing and other fields.
Nowadays, it is unfair for each of us to unknowingly contribute our chat data, shopping data, sports data, etc. to Internet companies for their use. Because they are using our data: first, we did not get our consent; second, we can’t ensure that the data is not abused; third, they have made money with our data, but not share with us.
This unequal situation should be adjusted in the next era. The data generated by each of us should have a calibrated value. With privacy-preserving computation, it will release the value of the entire then making data more valuable.
We also saw that most of today’s artificial intelligence companies face a very serious problem. Since data owners are reluctant to provide data and the companies do not have enough data to use, they have to work for the data owners. At the same time, artificial intelligence companies are also reluctant to contribute their core algorithms. This forms a thorough game relationship.
In the long run, a lot of human intelligence, whether it is data or algorithms, will be locked in the institutional server and it is difficult to make full use of it. We believe that a new generation of computing architecture based on MPC combined with VC architecture will change this phenomenon and create a new future.
We hope that everyone can join PlatON network in every way to release the value of data and build the next generation of infrastructure.
Thank you all.
