How blockchain becomes the TCP/IP protocol of finance, see what Xu Mingxing says!

In recent years, digital currencies, which were born in the Internet world, have attracted great attention from all walks of life. Digital currencies use cryptography and other computing technologies as a means to build decentralized trust through the construction of blockchains to achieve circulation and payment functions on the Internet.

On November 14, the "Closed-door Seminar on Global Blockchain Technology and Applications" hosted by the China Internet Finance Innovation Research Institute and the Financial Research Institute of the People's Bank of China was held in Beijing.

Leaders and experts attending the seminar included Wei Yingning, former vice chairman of the China Insurance Regulatory Commission, Xiao Feng, vice chairman of Wanxiang Group, Xu Mingxing, founder and CEO of OKCoin, and Qin Yi, partner of Deloitte.

What is blockchain? Simply put, blockchain, also known as distributed shared ledger, is one of the most disruptive Internet technologies in recent years. Blockchain builds a P2P self-organizing network, a time-ordered, tamper-proof ledger, and a distributed consensus algorithm to achieve decentralized trust. Currently, dozens of financial institutions including Nasdaq, NYSE, Citibank, UBS, Goldman Sachs, and Morgan Stanley are carrying out blockchain financial innovation. In addition to the financial industry, blockchain technology can also be applied to Internet business, government public information, electronic evidence, data security and other fields.

As a pioneer in the field of Bitcoin, Xu Mingxing, the founder and CEO of OKCoin, China's leading Bitcoin trading platform, shared his views on how blockchain can become the TCP/IP protocol of finance at the seminar. He believes that blockchain is a technology that can truly reconstruct the financial system from the bottom up.

Xu Mingxing likened blockchain technology to a toy:

“The early design of TCP/IP was actually designed by geeks and had no large-scale industrial applications. The same is true for today’s blockchain. It was originally designed by many hackers and is a very idealistic system.”

At this seminar, Xu Mingxing shared a lot of cutting-edge observations on blockchain. Let’s take a look at what he said. (The following is the full text of Xu Mingxing’s speech)

7 core technologies that form the foundation of the Internet ecosystem

To sum up, there are seven main parts: hardware, distributed relational database, distributed storage system, distributed large table, distributed memory database, distributed search engine, machine learning and deep machine learning.

First of all, there are hardware devices, such as terminals, cheap servers and network equipment. Distributed relational databases are a very critical technology for the Internet today. As we all know, Alipay achieved a peak of 100,000 transactions per second on "Double Eleven". This is not easy. This should be the highest record in the world. In the past, we all used Oracle database, which is the world's top hardware + software solution. But such a solution has long been unable to carry the huge volume of transactions on the Internet. Today, companies represented by Alibaba and Google, Alibaba is a distributed relational database, so there is "Double Eleven", otherwise there would be no "Double Eleven". Google has a very famous storage that can store massive amounts of data, which cannot be retrieved. It is a storage, just like our hard drive, which stores a bunch of things and needs to be found in a directory folder. In the past few years, there have been many tables in China, Big Table, which stores all the data in the world in a table. It is a distributed structure that can facilitate all data and the data structure of the entire big table. Therefore, whether it is Google or Baidu, such systems are all regularly rebuilding webpage systems, and they all retrieve data from the big table for retrieval. The latter is a distributed memory database. How can it be retrieved in real time? So there are new systems like this - distributed search engines, machine learning, and deep machine learning. These technologies make today's Internet so huge, which did not exist in the 56KB era.

Today's blockchain technology is also at the 28KB or 56KB model stage. How to promote it? Assuming that the system in China is, today's blockchain is absolutely unbearable. The current status of the largest Bitcoin system is that the upper limit of each block size is one megabyte. From 2009 to now, there are 380,000 blocks in the world, which store about 38GB of data, and the daily transaction peak is 300,000. There are nearly 10,000 (complete) nodes in the world, which is not that great, because many users are actually using Bitcoin service providers. For example, our wallet and trading platform serve 2 million users. This is the largest blockchain system in the world. The so-called distributed decentralized system and its current status are far from enough for industrialization.

Five major problems with the Bitcoin system

There are five major problems with the Bitcoin system: the total node scale is small and has not yet experienced a large-scale broadcast storm, transaction confirmation speed has slowed down, block synchronization speed is slow, the daily peak number of transaction processing is limited, and the system iteration update progress is slow.

First of all, because the early practitioners were geeks, and some were anarchists, the so-called purely decentralized systems they designed, because the total number of nodes in the Bitcoin system is still very small, and other systems have even fewer, there have not been large-scale broadcast storms, but recently someone did a stress test, and you will find that all Bitcoin wallet systems have collapsed.

The second is that the transaction confirmation speed will become slower and slower. With a large number of unconfirmed transactions piled up here, mining and sending a transaction may take ten hours to confirm. The current Bitcoin system has reached 38GB. For each terminal, it may take several days to use a Bitcoin wallet. This is unacceptable. The daily processing peak is 2,000 blocks, but it is 1 block every 10 minutes. Bitcoin developers are all very self-centered, so the Bitcoin system evolves very slowly. I often give an example that Christianity has forked, and Bitcoin is also going to fork now, Bitcoin 1 and Bitcoin 2. This is also a very prominent problem in the entire Bitcoin system.

Three development directions of blockchain: new public chain, private chain and side chain

There are three main directions for the development of blockchain: building new public chains, private chains and side chains.

Currently, Bitcoin is the world's largest and most widely used system. People are learning from Bitcoin technology and applying blockchain technology to other fields. There are currently several directions. The first is to build a new blockchain that has nothing to do with Bitcoin. A typical example is Ethereum. This blockchain is public and everyone can participate.

The other is private chain, which is familiar to everyone, such as R3 and Hyper ledger, which have realized distributed settlement on the blockchain and can be used by banks. For example, R3 talks to each bank and asks the bank to join the alliance and work together to form an alliance for research. Currently, more than 100 banks around the world have joined, such as Morgan Stanley.

The third is the side chain, which is very interesting. When Bitcoin's blockchain was first designed, it was just for me to be a trust, so the block structure did not support many complex data structures and operations. Blockstream put the starting point on a certain node of the blockchain, and Bitcoin's powerful computing power was used to maintain the authenticity of this set of nodes. Then I can define my own complex structure in the future, and can implement all the functions of Ethereum, but he can periodically hash his extended nodes and use Bitcoin's blockchain to assist in proving the credit of his public chain.

Engineering Design for Large-Scale Blockchain Networks

A star network that combines decentralization and centralization: a combination of full nodes and thin nodes, super nodes with massive processing and storage capabilities, transaction switches and routers, and a consensus mechanism.

First of all, we need to combine full nodes and thin nodes. Not every node data needs to run a data. For many people, it is OK to only store the data related to you.

Second, there must be super nodes with massive processing power. You can't ask everyone to be purely decentralized and all people are equal nodes. There should be some super nodes that can store all the data. It's no problem for us to store such data as super nodes. There are various data structures in the background to support it. This transaction cannot rely on pure broadcasting, otherwise it will be paralyzed. This is how the Internet came about.

Third, the switches and routers of Transaction must be referenced. The last one is the consensus mechanism. I think there are many misunderstandings, because many blockchain systems are currently referring to the Bitcoin system. However, Bitcoin's Proof of Work is not just for consensus. In fact, there are many ways to reach consensus. I think the entire large-scale blockchain network should be a new type of network that combines decentralization and centralization, similar to today's Internet. There must be regional switches and regional central nodes, so this should be a huge ecosystem, rather than running a node on everyone's computer. I can believe that Bitcoin's nodes will crash after running 10 million. These are some principles of large-scale blockchain network design.

How Internet Infrastructure Evolved

The evolution of Internet infrastructure requires several important elements: IEEE, W3C, continuous advancement of hardware technology, contributions from large companies, and talented technology geeks.

The early Internet is very similar to today's blockchain, but how did it evolve? It has several very important elements, such as IEEE (Institute of Electrical and Electronics Engineers), which is a very important organization. There is also W3C. The browsers we use today are all protocols made by W3C. Hardware technology needs to continue to break through. For example, many transactions can be made in the processor, and large companies have made a lot of contributions. In the early days of the Internet, a lot of open source software was made, but these industrial developments were continuously improved by Google and Microsoft. Android was originally made by one person, and then sold to Google. Now there is an Android ecosystem. Of course, this kind of genius geek is also needed. This person is the founder of Andy Rubin. The real blockchain investment and industry boom is far from coming. We should not say that this is our opportunity. We should let companies like Alibaba and Tencent enter this, so that the entire global large-scale blockchain ecosystem can be formed.

Possible technical misunderstandings in the development of blockchain

There are some problems that may exist in the development of blockchain at present. First, it solves one problem but brings more key problems. Second, the consensus mechanism is established on the basis of tradable digital currency.

I think there may be some problems in the current development of blockchain. The first problem is that many blockchain designs solve one problem and bring more problems. In fact, this is very typical in scientific research. Let me give you an example. I am studying superconducting physics. Superconductivity is a great discovery in the 20th century, and it is also the lifelong tragedy of many scientists in the 20th century, including my mentor and his wife. Both of them have white hair. The achievements are similar to those of decades ago. He realized the first generation of superconductivity, that is, low-temperature superconductivity. Superconductivity can only be achieved at a temperature of minus 260 to 70 degrees Celsius, but it is very rare. Superconductivity is impossible to achieve at high temperatures, so humans are desperate.

In the 1970s, these scientists were desperate because they could not industrialize. However, some stubborn people still studied all the materials on the earth. Finally, they turned their attention to ceramics, the best insulator. As a result, some superconducting materials were found in ceramics. They could achieve superconductivity at minus 200 degrees. Liquid nitrogen was enough. The cost was very low. It was no problem to run a liquid nitrogen system 24 hours a day. The whole world was boiling again. However, it was still impossible to use it. High-grade superconductivity is highly magnetic. This material has no plasticity and is all hard. It cannot be used to make wires. The physical mechanism of high-temperature superconductivity has not been figured out yet.

I give this example to say that just because you have achieved superconductivity, you cannot truly industrialize and apply it. It solves one problem but brings more problems. It can achieve superconductivity at the temperature of liquid nitrogen, but its material is ceramic, which is plastic and requires a lower temperature. Therefore, industrialization needs to consider the integrity of the system. The blockchain system needs to consider the overall performance and cannot only emphasize its advantages in one aspect.

Second, the consensus mechanism is based on tradable digital currency. I personally do not agree with this, because many blockchain systems want to design a new currency. Why? Because they refer to the Bitcoin system, and someone needs to mine, so that the so-called consensus mechanism can be implemented in the distributed system. But why do you encourage others to mine? I need a currency for him to mine, and he can get this currency. This currency can be traded and exchanged for money to form their motivation. However, in the thousands of years of human history, there are actually very few decentralized tradable systems, such as gold, silver, oil, iron ore, etc. Today's stocks and debts are dependent on the state. In the past few years, Bitcoin has emerged as a decentralized tradable asset. I know that Bitcoin trading is unreliable. It is an investment and speculation behavior. In the past, people from Ethereum asked if they could support the currency. I said that if we support you, we will destroy you, 50 cents a day, 1 yuan a day.

The central bank issues currency, including M0, M1, and M2. In fact, banks can add a token. China should have nearly 180 trillion in circulating currency, but it does not mean that all the money is printed. The printed money is very small, and a large part of it is in the bank's database. This token can be used in the blockchain network. For example, if we build a stock trading network today, we will use tokens for trading. After everyone exchanges tokens from the bank, decentralized settlement can be achieved in this network, and this network does not need to establish a consensus mechanism through mining. Moreover, this method is also opposed by the central bank around the world. The central bank thinks that it is OK for me to issue it. Therefore, there is no conflict between the real blockchain technology and the central bank's technology. We can completely decentralize the RMB 10 billion or 100 billion issued by the central bank in a blockchain network.

Possible future applications of blockchain

In the future, blockchain may be applied to many fields such as清算结算,股权交易,审计, and公证.

Blockchain is now being applied all over the world. One of the directions is clearing and settlement. In the past, large banks in the United States had to deposit money T+3. After the economic crisis in 2008, the Federal Reserve reduced the leverage of banks. Their liquidity became tight, and banks said they did not need this. Later, when liquidity became tight, they took advantage of Bitcoin to propose R3, and banks are slowly beginning to accept them.

The second is equity trading, which is done by Nasdaq. I think this is also a very promising thing. Today, the auditing companies and investment banks around the world have their value because of the opacity of our financial system in the past. Investment banks have many wealthy clients. If you want to sell stocks, I can help you sell them. How do investment banks sell? Go to the first LP and say that the company has already bought it, so you can also buy some. Go to another investor and say that the other five have already bought it, so you can also buy some. Investment banks have such capabilities. For example, if you want to do an IPO today, the transfer of your equity in the past and any person's subscription to your equity are all on this blockchain. This is a transparent system. You may not need investment banks to do these tasks in this transparent system. This is what Nasdaq did. It took 6 companies as pilots. I think this will have a very subversive impact on auditing and investment banks.

Factom is a notarization service provided by a US company. They provide proof of existence. When a person in the US buys a house, the ownership of the house may be stored in the database. In case of a dispute, has your database been tampered with? Did you collude with someone to change the property rights of your house? They register the land and property rights, and sign the database every 10 minutes or every 1 minute, and write it to the Bitcoin blockchain. I can prove that my property rights were not tampered with on a certain day in a certain month in 2013, because this is irreversible and is notarized by a global algorithm. This is a notarization service.

The Economist has an article that in a distributed system, where there is no trust, blockchain comes into play. For example, why do we have investment banks? It is because there is no trust in buying stocks, and auditing is also because of no trust. Why do clearing companies need to be used for clearing and settlement? It is also because of no trust that securities companies need clearing companies. I think we are already very advanced in the blockchain field, and we will not necessarily do worse than foreigners.