Prison Fantasia | Silk Road Founder Ross Ulbricht Releases Decentralized Community Platform "Proof of Concept"

The Silk Road is the most valuable e-commerce platform on the dark web, known as the "miracle of commerce," except when it comes to selling drugs. The founder of Silk Road, Ross Ulbricht, was sentenced to life imprisonment and is currently serving his sentence. While in prison, he proposed the concept of Decentralized Social Protocol (DSP). The original article "Decentralized Social Protocol" was published on the Medium blog, and this article provides a summary of its key points.

According to statistics from the National Development Council of Taiwan, Taiwanese people spend an average of seven hours online per day, with about two hours spent on social networking sites. This adds up to approximately 740 hours per year. However, centralized platforms have indeed encountered some issues, including privacy protection, custody, advertising, content filtering, and more.

The current "network effect" has led the original decentralization intentions of TCP/IP and HTTP protocols astray. Although content production remains in the hands of individuals, and individuals have the right to choose platforms, this is only a partial decentralization. Ross Ulbricht believes that true decentralization means that ownership and decision-making are controlled by users.

Advertisement - Please scroll down for more

Decentralization requires two things to be achieved:

1. Decentralization of Content Power: DSP allows creators to control who can access content through key-sharing encryption.
2. Business Model (Currency System): DSP service providers must create their own "business model," and in this system, users must be at the core of the value creation and transfer process.

Underlying Design Architecture

The current operation of centralized platforms consists of four components: Platform, Advertisers, Users, and Content Servers; the stakeholders are: Platform (red), Advertisers (green), Users (blue).

The operational logic is that the platform serves as the hub for the other three parties, controlling (storing) the content of creators, acting as a bridge between users and advertisers, and controlling the "total value" of the system.

However, in the Decentralized Social Protocol (DSP) architecture, users are the core, serving as the hub for platform service providers (Interface Providers), advertisers, and content servers. In this system, advertisers bid to place ads on the platform interface, pay users directly, and platform service providers and content servers compete for ad revenue.

In the decentralized social protocol, users can control all keys and control all value generated by the system.

Three Dimensions for Designing Community Platforms

So far, Ross Ulbricht has been discussing the differences in underlying operations. The next step is practical design.

Inventory of current community platform giants: Twitter focuses on short public statements, Facebook emphasizes interaction with friends, Reddit focuses on niche communities, Instagram is for photo sharing, and YouTube is a video platform; at this stage, many teams have begun to create decentralized versions of each platform, such as LBRY and Voice.

However, Ross Ulbricht does not need to replicate a decentralized version for each platform because these platforms are essentially the same, with differences only in the way they communicate and share content with others. By designing from this perspective, DSP can be simplified while covering various platform functions to the maximum extent.

He lists three dimensions:

1. Content Type: Although text, images, and videos may seem different, they can all be broken down into 0s and 1s. Therefore, platforms like Instagram, YouTube, and SoundCloud should be able to be designed on the same website.
2. Content Access: Content access is all about "permission issues." Ross Ulbricht believes that in a decentralized social protocol, encryption will be needed to ensure that only authorized individuals can view content, but it must be flexible enough for interface providers to design a variety of "content sharing mechanisms."
3. Contextual Dissemination Mechanism: Context matters; a joke can turn into a threat, and a troll can become a philosopher. Therefore, DSP must have a robust mechanism to understand context and treat it as metadata so that content can be presented according to the creator's intent. Here, context refers to other content related to the content, such as comments, likes or dislikes, shares, etc. The original meaning of context, in linguistics, refers to the meaning derived beyond the textual symbols due to culture, emotions, etc.

Interestingly, within DSP, there is no need for "content control." Through context, content dissemination can be restricted to reach the appropriate audience.

Three Specific Design Issues

After discussing the broader direction, Ross Ulbricht also mentions more specific designs, but he admits that these discussions are just food for thought, and he does not have the correct answers.

1. Personal Data Management

The main challenge for DSP in handling user data is the "namespace." Since DSP does not have a central list, it cannot check if registration names are duplicated.

Ross Ulbricht suggests using cryptography as a solution. The public key represents a user's identity in DSP, and the private key is kept by the client to prove that the person behind the identity is the user. Since public keys are pseudo-randomly generated, the probability of two people generating the same key is very small, almost negligible.

Alternatively, similar to reality, there may not be control over duplicate IDs, allowing multiple people to have the same ID. However, when exchanging documents, it is necessary to confirm the unique ID; another approach is to establish a name server mechanism, similar to how unique IP addresses are converted to unique domain names.

Or use blockchain to record non-duplicate keys or names. However, requiring registration on the blockchain would violate privacy because the blockchain is inherently public.

2. Establishing a Trust Network (Rating Mechanism)

DSP implements a rating mechanism to rate users, realizing decentralization to avoid the issue where each platform no longer commands the entire user base but retains a rating table liked or disliked by other users in the network and shares it with others. This concept is known as a Trust Network (WoT).

The real charm of WoT: there is no centralized perspective. If I label someone as a troll, those people are just trolls in my eyes, not necessarily the "truth." Some users may accept my judgment, others may ignore it, and some may consider the label as a positive evaluation.

WoT is not only used to filter out harmful content. It is also a decentralized system that helps users choose trustworthy business partners.

3. Value Creation and Transfer

Current centralized platforms have a successful business model, so to create a successful DSP, a dedicated business model for DSP needs to be redesigned.

The specific operation may involve advertisers bidding for positions on user pages, where users earn income from viewing ads on their screens. Users' clients sign ads with a password and send them to advertisers, allowing advertisers to know their ads have been viewed.

In this process, each step creates value and proves value, openly sharing this information to attract more advertisers. The more a user views ads, the more they can prove to other advertisers that they are a good investment, which allows them to earn more money. In a sense, this type of advertising is more targeted than what current platforms offer and will never share user content or violate their privacy.

When it comes to money, Ross Ulbricht also suggests that this structure should have a decentralized payment protocol, DPP, but it should be left to third-party developers to create plugins freely.

Now that users are earning money, the next question is:

What services should be provided to make users willing to pay?

Summarizing the four points mentioned by Ross Ulbricht in the article, these are the characteristics and potential situations he envisions for DSP:

1. Content Storage and Access

One of the services provided by centralized platforms is on-demand content storage and delivery. In DSP, this service also needs to be decentralized, and anyone must be able to provide this service at a very low threshold, while users must control how their data is handled. Since users need to pay for using the platform, and revenue comes from user-managed ad revenue, the more ads they accept, the more money they earn. This will also encourage content providers to compete, allowing users to access content faster and cheaper.

The operational model described in the article may seem unconventional. For example, when you search for a video, the search results will show you the prices and speeds of various content providers, and then you decide who to pay.

Additionally, the previously mentioned WoT can help the system run more smoothly. It will allow some content providers to establish a trust basis with users, enabling them to engage in lower-cost content exchanges with each other without using the bidding model mentioned above. This could evolve into content providers giving honors to loyal/important users or attracting new users to boost their trust rating.

2. User Interface

The "user experience" is also a crucial part of DSP. The user experience of social media mainly comes from dynamic messages, friend lists, message inboxes, homepages, chat windows, etc. These interfaces that shape the user experience have been controlled by centralized giants, such as the personal data you have on Facebook can only be seen on Facebook.

In DSP, any website or application can customize content and services according to user needs, and interface providers, such as websites or platforms, do not directly access content. This means that users "can" and "must" control the content they own.

It may be easier to imagine the examples in the article. When you open a page similar to Facebook, the left side is the friend list, the middle is the news feed, and the right side has some ads. In DSP, when you open the app, the business logic will generate this page according to the instructions of the interface provider, and other critical information such as profiles, private keys, and wallets are stored on the user's end. If you reinstall the app, it must request encrypted data from your content server, and users must authorize by entering a password to retrieve their original profile.

3. Extreme Scenarios

Under the design of users in DSP, using social media is profitable, but there may still be extreme scenarios:

In one extreme scenario, some users may not generate enough ad revenue to cover their costs. Another possible scenario is that if a user never watches ads, advertisers will eventually stop paying to display those ads. However, brand advertisers may still be willing to pay a small fee, allowing some content delivery.

Another extreme scenario is that there may be some users who are extremely valuable to advertisers, so much so that the ad revenue they bring exceeds the cost of using various DSPs.

The final scenario is "solitary" users who do not watch any ads to earn income, so they have to pay out of pocket.

4. Content Issues

While we expect DSP to be a censorship-free space, freedom is not without limits. Some content is harmful as it directly or indirectly causes harm to innocent people. A clear example is pedophilia. Weaker DSPs can become a tool for spreading such speech, leading to the failure of DSP. However, this poses a dilemma as open content can be reviewed but may harm privacy. Ross Ulbricht suggests a solution using Zero-Knowledge Artificial Neural Networks (ZKANN), utilizing cryptography and network machine learning to achieve this. Ultimately, however, something similar to a censorship algorithm must be developed to address content issues.

Conclusion

Ross Ulbricht believes that for DSP to be successful, it must seamlessly integrate experiences but also have "extra benefits" that centralization lacks, including encryption, content servers, micropayments, ZKANN, and even users being able to earn a little money by viewing ads.

However, Ross Ulbricht also admits that he has not researched the current status and latest developments of DSP as he is still confined in prison. This article is mainly to stimulate discussion and hopes to assist those researching DSP.