Analysis: Why did Twitter founder Jack Dorsey develop protocols instead of platforms?

Author: Wang Guangzhong

Editor's Note: The original title was "Negotiate Agreements, No Platforms"

On December 11, 2019, Twitter founder Jack Dorsey tweeted that he would invest in setting up an independent open standards R & D team to develop a set of protocols for social applications. The ultimate goal is to make Twitter a client of this set of protocols. The tweet did not mention the word Web3.0, but only made it clear that Twitter was going to switch from "platform-based" applications to "protocol-based" applications. But this is exactly the core content of Web3.0. In addition, Jack also mentioned explicitly that blockchain will be used, so Twitter's direction is obviously highly consistent with Web3.0's thinking. The title of this article, "Agreement over Platform," comes from the title of an article mentioned in the tweet [1].

So what benefits would an Internet application like Twitter develop to a "protocol-based" approach for companies or users? Is it really feasible in reality? Lets come look.

Protocols and platforms

Currently, some of our daily applications are protocol-based applications, but the mainstream is platform-based applications. Let's first introduce the differences between protocols and platforms.

The so-called agreement is an open game rule, which is open and neutral. For example, we use the Web through the HTTP protocol. This protocol is released by a non-commercial standardization organization. Anyone using the technology involved in the protocol does not need to pay a fee. There are some clear rules in the protocol. For example, when using the HTTP protocol, if you enter http://example.com/about.html in the browser address bar. The stipulation in the HTTP protocol is to first find the server corresponding to example.com, and then download the about.html page to a browser and open it. Open game rules mean that anyone can participate. For example, based on the HTTP protocol, any manufacturer can develop a browser. Any browser developed by anyone can join the network without discrimination as long as it complies with the HTTP game rules. Coming. There are many examples of Internet protocols. Chat has the IRC protocol, video communication has the WebRTC, and Email has the SMTP protocol.

The so-called platform is a private company. The "platform-based" application is characterized by the non-publicity of rules. For example, Facebook is a typical platform. How the application works internally, how it communicates, and how the underlying format of the data is determined by their company. In this way, the data is bound to the application, and the company owns the user's data. Because of the data on Facebook, users cannot download it completely, even if it can be downloaded, they cannot be used in other applications because the data format is different. Web2.0 is also called "Platform Internet" because mainstream applications such as Amazon, Facebook, and Twitter are "platform-based."

In short, on the surface, the agreement is a document, and the platform is a company. There seems to be no comparability, but in essence the agreement represents an open game rule, and the platform is a set of closed game rules.

Advantages and disadvantages of the platform

Jack also mentioned in his tweet that Twitter was very similar to a set of protocols in the early years, but then it took another path and gradually became centralized and developed into a platform. This is not alone. Many "protocol-based" applications in the early days of the Internet did not really take off, but were replaced by platforms. For example, "protocol-based" Usenet was replaced by platforms such as Reddit. So what are the advantages of the platform over the protocol? Now that Twitter has decided to take the "protocol-based" route, what are the disadvantages of the platform?

In the Web2.0 era, the platform wins, not without reason. Compared with the protocol, the platform has many advantages. "Protocol-based" network applications have many obvious problems. The first is that it is not easy to upgrade. Everyone has their own client. To upgrade the agreement, they need to reach a consensus. The well-known case is the "browser war." Second, because the rules are fair, effective control cannot be achieved. If the boss does not have any privileges, it will be very difficult for the boss to appear. Without the boss, no one can settle for many things. Third, it is difficult to commercialize because the rules are open. A prerequisite for commercialization is the definition of scarce resources. Open rules have no boundaries and are difficult to define. Corresponding to these problems, the platform has obvious advantages. First of all, if you want to upgrade your company's application, you can do it immediately. Second, the application belongs to the company, and effective control is easy, for example, it can effectively review bad content, optimize user experience, and so on. Third, because of strong control, scarcity can be defined at will, so commercialization is easy to achieve, such as by selling membership fees. In addition, at a deeper level, the platform provides the trust necessary for business activities. Imagine if we did not trust Alipay, how could we secure a deal.

But over time, problems with the platform have also been exposed. First, the platform is subject to legal pressure. Take social platforms as an example. At the same time, regulators will think that the platform review is too loose, and users will feel that the review is too strict, and they may be blocked under almost complete unconsciousness. You know, a global social platform must adapt to regulators and users with completely different cultural backgrounds, so the development of the platform is difficult. Secondly, the trust crisis of the platform is also getting worse. Facebook's privacy leak scandal has made Zuckerberg's hearings commonplace. Third, the platform has private business interests in mind, and interest driving can cause direction distortion. The content platform was originally pursuing the highest content quality, but now the homepages of many content platforms are full of vulgar titles.

In short, the advantage of the platform comes from the strong control and trust brought by centralization, but today, the biggest problem of the platform also lies in centralization.

Advantages of the agreement

Therefore, it is more reasonable for companies like Twitter to switch back to a decentralized state. So what are the advantages of "protocol-based" applications?

First, "protocol-based" allows decision-making power from the center to the edge of the network. What content to recommend on the platform is not determined by the centralized company, but by each user at the edge of the network. Users can choose their own client or install their favorite recommendation algorithm plugin to decide what they want to see. For example, if someone hates the topic of homosexuality, then you can block them by setting it yourself, without forcing everyone to watch related topics. In this way, the benefit is that Internet companies have no legal burden and can write software with peace of mind. At the same time, users will feel more free to publish content and will be more active on the platform.

Secondly, "protocol-based" can return data control to the user. Users control their own data for better privacy background security. Data is not held by a single platform company, which also gives startups more opportunities. For example, before Gmail, Hotmail and Yahoo Mail have been very popular, but it does not prevent Google people from winning users with a good user experience. Google can analyze the data stored in Gmail to advertise to customers. This seems similar to Facebook, but it is essentially different. Because the data in Gmail can be exported in full, whether it is contacts or emails. If Gmail is very unsatisfactory for users, it is easy for users to bring the data and change to an Email client. But Facebook data, even if it can be exported, can't guarantee that I and my friends will continue to contact.

Of course, for the "protocol-based" thinking to really rise, it needs to be commercially viable. "Protocol-based" applications can make money by recommending value-added services such as acceleration or storage. For example, the Gmail client is free as an application. However, users also tend to use Google's storage service to store their own mail. Data, Google can then make money by parsing the data and pushing some ads. Commercialization is another big topic, so I won't start here. It needs to be emphasized that the defects of the previous "protocol-based", such as the problem of effective control and difficult commercialization, have been solved with the birth of the blockchain.

Jack's tweets sum up the advantages of "protocol-based" application development that we mentioned earlier. First, Jack found that some things are difficult to achieve with a centralized platform, for example, a set of global Second, social application recommendation algorithms are not open source, and users have no freedom of choice. Third, social applications, for their own benefit, will promote content that is easy to attract attention, not healthy content. Fourth, new technologies make decentralization possible.Blockchain provides an open and sustainable data storage strategy, as well as governance and monetization mechanisms.

in conclusion

To conclude, give answers to the questions at the beginning of this section. "Protocol-based" can make applications decentralized. The platform can avoid the regulatory dilemma and focus on developing high-quality software, and users get more freedom to publish content and flexibility in filtering content. However, the vision of "protocol-based" development wants to be truly implemented, and to address the various weaknesses brought by the openness of the protocol, such as ineffective governance and commercialization. Blockchain is the key technology to overcome these weaknesses.

Also, Twitter is not the first company to develop a social protocol. The father of the Web, Tim Berners-Lee's Solid Project [2], is the exploration in this direction.

[1] https://knightcolumbia.org/content/protocols-not-platforms-a-technological-approach-to-free-speech [2] https://inrupt.com/solid