Future of Social Networks (2 of 3)

*Forward the Original Title：Future of Social Networks (2 of 3)

In 2017, a group of MIT Media Lab researchers claimed in Wired that decentralized social networks “will never work” [1]. In their piece, they cited three impossible challenges: (1) the question of onboarding (and retaining) users from scratch, (2) the (mis)handling of personal information of users, and (3) lucrative user-targeted user advertisements. In all three cases, they argued, incumbent tech giants, such as Facebook, Twitter, and Google, simply had too far-reaching economies of scale to make room for any significant competition.

Fast forward half a decade later, what was once hailed as “impossible” seems no longer so far-fetched, and we seem to be on the dawn of a paradigm shift in the way we conceptualize social media networks. In this three-part series, we will examine how new ideas in decentralized social (DeSo) seem to address these “age-old” questions, specifically, (1) the use of open social graphs in solving the cold start problem, (2) using proof-of-personhood and cryptographic techniques to solve the userhood problem, and (3) leveraging tokenomics models and incentive structures to solve the revenue problem.

Social Media’s Userhood Problem

Modern social media suffers uniquely from a bot problem. Although social media platforms have a mandate to uphold freedom of expression, this issue becomes thorny when the “users” in question are not in fact real users, but bots. And, as it turns out bots can have a significant impact on public discourse, from alleged tampering with US presidential elections to influencing the public opinion on COVID [1]. Especially with its emphasis on anonymity, security, and privacy, any decentralized social media platform inherits the “bot problem” – essentially, how do you convince people that the accounts on your platform are real and not bots, especially in an era of advanced AI?

The naive approach is simply just a traditional know-your-customer protocol, but this approach immediately runs into a privacy problem – the flip side of the coin. How (and why) should you trust any social media platform to hold a treasure trove of our sensitive data (from government IDs to private messages and financial transactions) that is able to recreate someone’s entire personal, social, and professional life?

Thus, the “userhood” problem is a tension between confirming that users are actually human versus making privacy guarantees on personal data. Within this writing, we will explore two distinct approaches to tackling this problem, a biometric approach (with zero-knowledge proofs) and a social-vouching approach.

Worldcoin and Biometric Authentication

Within the “proof-of-personhood” problem space, Worldcoin stands out as one of the most notable and controversial projects. In addition to having Sam Altman, the famed CEO of OpenAI as one of its proponents, Worldcoin’s solution to the “proof-of-personhood” question is very straightforward: use a retina scan to create biometric proof that you are a human (since bots don’t have retinas yet), and receive an authentication token from this. As for data privacy, Worldcoin claims to use Zero Knowledge Proofs to ensure that the biometric data obtained is stored securely [2].

Worldcoin Orb. Image Source: https://www.wired.com/story/sam-altman-orb-worldcoin-tools-for-humanity/ [3]

The thesis behind Worldcoin is that with the increasing role that AI plays in society, there needs to be a way of telling humans and bots apart, crucially in a privacy-preserving and decentralized way. Through using the Worldcoin orbs’ retina scans, one can obtain a “digital-passport-like” World ID, which enables recipients to potentially be eligible for a crypto-based Universal Basic Income mechanism and participate in novel mechanisms of global democratic governance [3]. In essence, this World ID is intended as a social primitive to bootstrap the digital social networks of the future.

Throughout its documentation, Worldcoin emphasizes how it has a privacy-first solution. For example, it states that it deletes images collected by the Orb, storing only a hash of the user’s iris, and runs Zero Knowledge Proofs (zk-SNARKs) in order to share proof of personhood information without disclosing any personal data. And although in the current phase of rolling out, these hashes are stored in a centralized database, the team is dedicated in the long-run to store these iris hash data on-chain after the hashing algorithm is fully mature [4].

But despite these privacy-preserving claims, there are still numerous controversies about the true privacy, security, and fairness guarantees. For example, there have been claims of Worldcoin operators having their credentials stolen, and World IDs being sold on digital black markets, such that users could attain Worldcoin tokens without going through the iris scan themselves [5] [6]. There also have been overall equity concerns, with the MIT Technology Review publishing a scathing article in April 2022 on deception, manipulation, and exploitation of nearly half a million users (primarily in developing countries) during its test phase, going so far as to call it a form of “crypto-colonialism” [7]. Indeed, as of August 2, 2023, Kenya, formerly one of Worldcoin’s largest collection venues, had banned Worldcoin scans over security, privacy, and financial concerns [8].

Apart from these project-specific controversies though, there also exist broader concerns about Worldcoin’s overall approach of biometric authentication through dedicated hardware. Because the Orb is fundamentally a hardware device, even if Worldcoin’s software was all perfect, there is no way to guarantee that there is a hardware backdoor that allows Worldcoin (or another third-party manufacturer) to secretly collect users’ actual biometric data, or insert fake profiles into the system [9]. To skeptics, it can appear that all of Worldcoin’s privacy assurances (ZKPs, iris hashes, decentralization on-chain) seem to be no more than an ironic statement of “trust me bro, we’re a trustless solution.”

Proof of Humanity and Social Vouching

A different approach to the proof-of-personhood problem is to use a social-vouching approach. Essentially, if verified humans Alice, Bob, Charlie, David all “vouch” that Emily is a verified human, then chances are that Emily is probably also a human. The core question here is therefore a question of game theory design – how do we engineer incentives in a way that maximizes our ability to “verify humans.”

From Proof of Humanity Website

Proof of Humanity is one of the oldest and most important projects within this space. In order to “prove your humanity,” you need to (1) submit your personal information, photos and a video, as well as a 0.125 ETH deposit, (2) have humans already existing in the registry vouching for you, and (3) pass before a “3-challenge period.” If anyone challenges you during this period, this case is brought to the Kleros decentralized court, with this deposit at stake [9].

Within the vouching process, the user is first paired with a voucher through a spreadsheet of vouchers. After the user pairs with their voucher, they then do a video call in order to validate that the profile matches with the real person [10]. Like the thesis of Worldcoin, the Proof of Humanity community has in the long run an idea of Universal Basic Income (UBI) in mind, available to those verified within the Proof of Humanity registry [11]

Some other projects following a similar path of leveraging social graphs to authenticate personhood include BrightID’s video call verification, where everyone verifies each other, Idena’s continuous captcha creation and solving games, and Circles’ trust-based cliques.

Perhaps the biggest appeal of these social-vouching based platforms is that they don’t seem to be as intrusive as Worldcoin, which literally requires you to scan your iris at a metal orb. Some of these approaches, such as Idena’s captcha “checkpoint rituals,” even seem to preserve some degree of anonymity, not requiring large amounts of personal data sharing or needing a third-party identification center [12].

The Future for Proof-of-Personhood

As AI continues to advance and exhibit increasingly human-like behavior, it is increasingly important to devise novel mechanisms for proof-of-personhood, not only for Universal Basic Income and other incentives that many of these proof-of-personhood projects discuss, but more importantly as a way to better sanitize and regulate the social networks of the future.

Yet, from data privacy to invasiveness of process to effectiveness at determining personhood, this process is one that involves numerous tradeoffs, being one of the famed “hard problems in cryptocurrency” [13]. As Vitalik himself notes, there does not seem to be a single most ideal form of proof-of-personhood, and puts forward a possible hybrid path as a suggestion: one that bootstraps using biometric-based approaches, but in the long-run transitions to more social graph-based approach.

Biometric-Social Graph Hybrid Path [9].

Nonetheless, going forward, this is a space that requires much more transparency of process, code, and data. In short, there cannot be an ironic paradox where users need to “trust that it’s a trustless solution”. It is only through this way that we can truly create a social network primitive that stands true to crypto’s original vision of decentralization and privacy.

Disclaimer：

1. This article is reprinted from [[VeradiVerdict]. Forward the Original Title‘Future of Social Networks (2 of 3)’.All copyrights belong to the original author [*Paul Veradittakit]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.