Vitalik's latest paper: Decentralized Society - Finding the Soul of Web3 (Part 2)
Authors: E. Glen Weyl, Puja Ohlhaver, Vitalik Buterin
"The Taoist is the mystery of all things, the treasure of the good man, and the protection of the bad man"
——Chapter 62 of Laozi
Summary
Authors: E. Glen Weyl, Puja Ohlhaver, Vitalik Buterin
"The Taoist is the mystery of all things, the treasure of the good man, and the protection of the bad man"——Chapter 62 of Laozi
SummaryPlease check the previous article:
7. Challenges in Implementation
Privacy concerns are a key challenge for DeSoc. On the one hand, too many public SBTs may reveal too much information about souls, which makes them passive and subject to "social control". On the other hand, too many purely private SBTs will also lead to the problem of discounting the correlation between private communication channels and social governance and coordination, which reflects the importance of incentive compatibility. Also closely related to the issue of privacy is the issue of deceit: spirits may communicate through private or other secondary channels, thereby distorting their communal solidarity. It is impossible for us to know all the possibilities and answers along the way, so we need to deeply explore the nature of these difficulties and draw a promising path for the future.
7.1 Private Souls
The blockchain system is open by default, and any relationship recorded on the chain is immediately visible not only to the participants, but also to anyone in the world. Having multiple aliases allows for some privacy: a family soul, a health soul, a professional soul, a political soul, each corresponding to different SBTs. But if these aliases are superficial, outsiders can easily connect these souls together, and the consequences of this behavior are serious. So, if no measures are taken to protect privacy, the "naive" act of simply putting all SBTs on-chain will result in a large amount of personal information being exposed on a crowd of applications.
secondary title
"Game" is an important problem, and solving this problem is one of the priorities of future research. In fact, this is one of the main reasons why it is very difficult to open source existing algorithms that provide prioritization or ordering. In order to reduce and stop the SBT "game", we offer several norms and directions:
It is up to the individual to choose how to store off-chain data. Possible solutions include (i) own devices, (ii) trusted cloud services, and (iii) decentralized networks such as the InterPlanetary File System (IPFS). Storing the data off-chain allows us to have the permission to write SBT data in the smart contract and have separate permission to read the data. Bob can choose to display the contents of any of his SBTs (or other data it stores) only if he wishes. This is a big step forward, and since most data needs to be processed by only a few people, it further improves the scalability of the technology. However, to fully realize the protection of multiple privacy features (annotation: refers to the collection of various types of privacy or (and) privacy), it is necessary to dig into the relationship in more detail. Fortunately, many encryption techniques can help us with this.
Now there is a powerful set of buildable blocks that support new ways to partially reveal data information, it is called "Zero Knowledge Proofs" (Zero Knowledge Proofs), which is a branch of cryptography. While zero-knowledge proofs are commonly used today for privacy protection of asset transfers, they also allow one to prove arbitrary statements without revealing any information other than the statement itself. For example, in a world where government documents and other credentialing information can be cryptographically proven, someone could attest to statements like "I am a Canadian citizen, over 18 years old, have a degree in economics from a university, have over 50,000 followers on Twitter, and someone You have not registered an account in this system yet."
Zero-knowledge proofs can be computed on SBTs to prove properties about a soul (eg it has certain members). This technology can be further expanded by introducing multi-party computing technology (such as garbled circuit computing), which makes the proof process have two-way double privacy: the verifying parties do not disclose who they are, and the verifying parties do not disclose their verification mechanism. In this process, both parties calculate together, and only information is output.
Another technique is to specify validator proofs. In general, "data" is unreliable: if I send you a movie, I can't technically prevent you from recording it and sending it to a third party. Methods like digital rights management (DRM) are limited and often cost the user a lot. But the "proof" is solid in a way, if Amma wants to prove to Bob some property X of her SBTs, she can do a zero-knowledge proof of the following statement: "I hold SBTs satisfying property X, or I Possesses an access key to Bob's soul." bob is convinced by this statement: Because he knows he didn't make a proof, Amma must actually hold SBTs that satisfy property X, but if bob passes the proof to Cuifen, Cuifen doesn't Will be convinced: because as far as he knows, bob can prove it with his own soul access key. At this point the proof can be further strengthened using Verifiable Delay Functions (VDFs): Amma can show a proof that can only be produced now with the required SBTs, but others will have to wait until 5 minutes later. This means that access to a trusted proof of the data is possible, although the different types of raw data itself (possibly copied and pasted) cannot be selected. Just as traceability in blockchain transactions prevents someone from copying and pasting a valuable NFT (and Sybil attack on the original sender), SBTs can provide traceability in propagation, which at least reduces the risk of uncertain origin. Value of data (copy-pasted).
These off-chain data and zero-knowledge techniques are compatible with negative reputations (embodied by SBTs), which will still be revealed even if the holder does not want them to be seen. Negative reputation includes credit history, outstanding loan data, negative comments and complaints from business partners, and the degree of harmony with relevant social relations as evidenced by SBTs. The combination of blockchain and related cryptography could lead to a potential solution: smart contracts could force souls to incorporate negative SBTs into a data structure such as a Merkle tree stored off-chain, any zero-knowledge proof or Garbled circuit calculations all need to incorporate this information, otherwise, there will be a visible "gap" in the provided data and the verifier will be identified. The Unirep protocol is an example.
The point of these examples is not to show how cryptography can be used to address all privacy and data permission concerns of SBTs. Rather, it outlines a few examples that demonstrate the power of these technologies. An important future research direction is to identify the boundaries between different types of data permissions, and the specific combination of technologies that are best suited to achieve the desired level of permissions. Another question is what type of composite property regime is required for data governance and how to separate use rights (“usus”), construction rights (“abusus”), and usufruct rights (“fructus”).
7.2 Cheating Souls
If SBTs are the social basis for coordinating composite property, networked goods, and intellect, there is a concern that spirits may gain access to communities through trickery or deceit to gain governance or property rights that SBTs license. For example, if many application directions depend on SBTs that can represent the right to attend meetings, then it may be possible to use these SBTs in exchange for bribes. Bribing enough people, humans (and bots) generate a fake social picture separated by (fake) SBTs. Just like DAOs can be bribed, so can souls and the on-chain voting mechanism they use. Conversely, the impact of SBTs can be mitigated if they are used to weaken collaboration. Why should we trust that SBTs possessed by spirits actually keep their social commitments, rather than simply telling them how to play this "game"?
One view is that there is a "balance" between the different motivations for deception. Souls self-measure and classify the networks they deem important, much like how Harberger taxes work to arrive at near-accurate market valuations after balancing the incentives to overvalue and undervalue assets. Souls will want to have more SBTs to gain influence in their communities, on the other hand they will avoid SBTs from communities they don't care about and thus score lower on relevant metrics, thereby enhancing their wider influence in network governance.
But it would be naive to think that the two motives of gaining power and gaining influence will always cancel each other out (or come close to canceling out). There may be many communities using systems other than SBTs to limit access and governance. Alternatively, the community might issue private SBTs (contrary to our assumption about publicity) to bolster governance power, while inducing community members to keep the existence of these private SBTs secret from broader decision-making.
"Game" is an important problem, and solving this problem is one of the priorities of future research. In fact, this is one of the main reasons why it is very difficult to open source existing algorithms that provide prioritization or ordering. In order to reduce and stop the SBT "game", we offer several norms and directions:
The ecology of SBTs can be opened from "intensive" social channels, where SBTs confirm off-chain community membership through strong social bonds and interactions with each other. This makes it easier for the community to identify, filter and revoke fake (or bot) SBTs. We often find such "dense" channels in churches, workplaces, schools, congregation groups, and civil society organizations, which would provide a "cop game" played in more "sparse" social channels (e.g. via bots, bribes, etc.) , impersonation) provide a more Sybil-resistant social foundation.
Nested communities require SBTs to impose "context" on their "downward" potential collusion vector. For example, if a state is holding a fundraising or ballot round, the state may require that every participating citizen also hold an SBT for the designated county and city.
The openness and cryptographically provable nature of the SBT ecosystem can be used to actively detect collusion patterns and punish unreliable malicious behavior (maybe reducing the voting weight of colluding souls, or forcing souls to accept SBT - which in this case represents negative reputation). For example, if a soul proves that another soul is a robot, the case can drill down and publicly verify the results, resulting in a substantial negative reputation for that soul to prove. A similar use case has arisen in the GitCoin QF ecosystem, which uses a series of indicators or signals to detect "colluding groups".
Zero-knowledge proof technology (such as MACI) can prevent certain proofs made by souls from being provable through encryption. This discourages the sale of proofs, since the briber cannot tell whether the briber fulfilled the transaction. There is already a great deal of research on this technology, and eventually any non-financialized social mechanism could benefit from similar ideas.
The ecology of SBTs can be opened from "intensive" social channels, where SBTs confirm off-chain community membership through strong social bonds and interactions with each other. This makes it easier for the community to identify, filter and revoke fake (or bot) SBTs. We often find such "dense" channels in churches, workplaces, schools, congregation groups, and civil society organizations, which would provide a "cop game" played in more "sparse" social channels (e.g. via bots, bribes, etc.) , impersonation) provide a more Sybil-resistant social foundation.
Nested communities require SBTs to impose "context" on their "downward" potential collusion vector. For example, if a state is holding a fundraising or ballot round, the state may require that every participating citizen also hold an SBT for the designated county and city.
Zero-knowledge proof technology (such as MACI) can prevent certain proofs made by souls from being provable through encryption. This discourages the sale of proofs, since the briber cannot tell whether the briber fulfilled the transaction. There is already a great deal of research on this technology, and eventually any non-financialized social mechanism could benefit from similar ideas.
We can encourage whistleblowing behavior, which destabilizes mass "collusion". This is not the detection and punishment of incorrect or abusive behaviour, but the complicity of patterns of abuse. Overuse of this technique is risky due to the possibility of bogus bribes, but it is still inherently viable.
If some souls have a common interest, we can use a correlation coefficient which measures the correlation. For example, use correlation techniques from quadratic financing to quantify the correlation between two players and thereby determine their degree of intersection. If two players have many interests in common, their incentive to reveal this fact (there are many interests in common) to the quadratic funding mechanism will certainly diminish with the relevance discount, but it will never become zero or negative .
first level title
8. COMPARISONS AND LIMITATIONS
While the scope of proposed identity frameworks is virtually limitless, there are four prominent and close paradigms in web3 that deserve comparison: the dominant "legacy" authentication system, the pseudonym economy, proof of personality, and verifiable credentials. Each paradigm highlights important contributions and challenges to future development of the social identity paradigm we advocate, and we use these limitations as a springboard to explore future directions. Taken together, we also explain why we believe souls representing social identities and soul-bound tokens are a more promising direction for privacy regimes.
8.1 Legacy
Legacy authentication systems rely on documents or IDs issued by a third party (government, university, employer, etc.) through which provenance is determined. While legacy systems deserve deep understanding, these systems are highly inefficient and lack composability to enable fast, effective coordination. Additionally, these systems lack the context of social relationships, making souls dependent on a centralized third party to affirm community membership, rather than being embedded in it.
For example, most government-issued ID cards are ultimately traced back to birth certificates issued by doctors and family members, who are the ultimate source of truth, but this also ignores many equally meaningful social ties that combine to , which provides a strong validation.
In fact, when centers of concentrated power need to seek strong buy-in (such as obtaining security clearances from governments), they rarely rely on these documents, turning instead to the “social relationship” route. As a result, such legacy identity systems tend to concentrate power on the issuer and those who can do “due diligence” to obtain stronger proofs, which in turn create rigid and unreliable bureaucracies. A key goal of DeSoc is to ensure that government ID security requirements are met and exceeded, allowing horizontal networks to provide greater security for all users through a range of social foundations.
secondary title
8.2 Pseudonymous Economy
It envisions people accumulating transferable zero-knowledge proofs in their own wallets and evading reputation attacks by splitting the proofs into new wallets or multiple wallets, possibly without traceability. When picking proofs to transfer, there is a trade-off between the degree of pseudonymity required for the new account, which requires a choice between being more anonymous (less proofs transferred) or distributed to the social network (more proofs transferred).
The practical difference between the typical pseudonymous economy proposal and DeSoc is that we de-emphasize that segregation of identities is the primary means of protection from a culture of "accusation." Some degree of segregation (e.g. different souls between family, work, politics, etc.) can be beneficial, but in general, relying on new identities as the primary defense against attack is highly detrimental and it makes lending and provenance Reputation staking becomes difficult, and it composes poorly with governance mechanisms that try to correct correlation or Sybil attacks.
DeSoc does not by allowing the victim to re-emerge in the attack with a new identity (if diminished), but allows other methods such as socializing the attacker. “Allegations” come up often because when a person (or bot) has little social connection to the victim, the statement and action are disengaged, and the slanderous information spreads through non-relational networks. In the same way that SBT provides attribution to prevent falsification, SBT traces the attribution of "slanderous behavior" on social connections. “Slandering behavior” is essentially the product of being outside the victim community (as reflected by shared SBT members), or lacking proof of SBT from the victim community (which casts doubt on the authenticity of the behavior).
SBT also enables victims to mount defensive responses to counteract blows engineered and propagated from their trust network (represented here by the co-owned SBT model). By maintaining social relationships, people can maintain trust even when they face the threat of "accusations" and hold attackers accountable. Improving provenance improves the social basis of truth.
8.3 Proof of Personality (PoP)
The Proof of Personality Protocol (PoP) aims to provide personally unique tokens to prevent Sybil attacks and allow non-financialization of applications. To do this, they rely on methods such as global analysis of social graphs, biometrics, synchronized global key players, or some combination thereof. However, since the PoP protocol seeks to represent individual identities (to achieve global uniqueness), rather than mapping relationships and solidarity social relations, the core of the PoP protocol is to treat everyone equally, and most of the application directions we are interested in (such as reputation pledge) , are all about being human and transcending being a "different" person to being a "unique" person.
Also, PoP protocols are not immune to Sybil attacks. In almost all near-term foreseeable applications, PoP systems are vulnerable to Sybil attacks, albeit at a slightly higher cost. Unless the majority of people on the planet have signed up for a PoP service and participated in a specific verification activity, an attacker can always recruit non-participating (or uninterested) people to act as "witches". While not all bots are hired, there is little difference, except maybe some added fees.
Many PoP agreements aim to build a foundation for universal basic income or global democracy. Although we do not have the same ambitions, these agreements prompt us to think about how to gradually build and coordinate multiple network products. Unlike the binary, individualistic, and global nature of PoP, we aim to build a rich, layered, and interconnected foundation for bottom-up reputation, property, and governance, and allow for participation in communities and networks of all sizes .
8.4 Verifiable credentials
Verifiable Credentials (VCs) are a W3C standard whereby credentials (or certificates) can be shared at the holder's discretion with zero knowledge. VCs highlight key limitations of our baseline privacy paradigm and motivate us to explore further the aforementioned privacy content. VCs and SBTs can be seen as natural complementary elements until SBTs have the function of narrowing the scope of opening: in particular, SBTs are initially public, so they are not suitable for sensitive information such as government-issued identification, while the implementation of VCs has been In grappling with a recovery paradigm, this could be addressed by community resilience. In the short term, the combination of the two approaches works better than either approach alone. But VCs have a key limitation: At least in general, VCs do not support most of the application directions we enumerated because of their unilateral privacy.
Unilateral ZKC is not compatible with our use case, nor does it meet our normative definition of privacy. Most of our application directions rely on some level of publicity, but with zero knowledge sharing, a soul cannot know that another soul has an SBT unless it is shared with the other party. This makes reputation pledges, credible commitments, anti-sybil governance, and simple rental contracts (such as apartment leases) impossible to have visible other commitments or proof of title.
first level title
9. SOUL BIRTH
The road from the current web3 ecology to the enhanced sociality of SBTs faces a typical cold start challenge. On the one hand, SBT is non-transferable, on the other hand, the current wallet form may not be the ultimate destination for SBTs, because they lack community resilience mechanisms. But in order for community resilience wallets to work, they need to provide different SBTs in a decentralized community to ensure security.
secondary title
9.1 Initial SBTs (Proto SBTs)
While SBTs are non-transferable, there may be another attribute of SBTs that will play a role in development: revocability. SBT can first become a revocable, transferable token before growing to be non-transferable. Tokens are revocable if the issuer can burn the token and reissue it to a new wallet. For example, when keys are lost or compromised, and the issuer is interested in ensuring that the tokens are not monetized and sold to one party. (In other words, burning and reissuing will make sense when tokens signify true community membership.) Employers, churches, meetup groups, off-chain interactive clubs with multiple exchanges are all ways to burn and reissue tokens. A good place for a coin because they have a relationship with someone who can easily check for an imposter through a phone call, video conference, or simple face-to-face meeting. A single interaction, such as attending a concert or conference, is not suitable because of weak community connection.
Revocable and transferable are the initial characteristics of the initial SBT before the birth of the soul. These tokens buy time for wallets to foster a secure community resilience mechanism and for individuals to accumulate initial SBT that can eventually be burned and reissued into non-transferable SBT. Under this approach, the question is no longer "SBTs or community resilience first?" Instead, SBTs and community resilience mechanisms work together to give birth to a soul.
9.2 Community Recovery Wallets
While today's wallets lack community resilience, they each have strengths and weaknesses as home or "breeding places" for SBTs. The Proof of Personality (PoP) protocol has the advantage that it is already trying to build a social dispute resolution mechanism, which is the foundation of community resilience. Additionally, many DAOs use POP to facilitate governance, making them a natural number one issuer of SBTs. However, despite PoPs being at the forefront, PoP protocols have yet to gain widespread trust for storing valuable token assets, whereas custodial wallets have.
Hosted wallets (heavy centralization) are therefore dependent on immature users. Such custodial wallets could build tools for the retail community to issue revocable tokens that can then be converted (or reissued) into SBT, and even for more "enterprise" issuers who lack relevant expertise ( Many of them seek to build a foundation of loyal customers in web3). Once the community resilience mechanisms are formalized and tested, these escrow wallets can be decentralized into community resilience, while custodians continue to provide other valuable services at DeSoc (e.g. community management, issuance of SBTs, etc.).
secondary title
9.3 Proto-Souls
Codes of conduct can also guide the soul into existence. As we rethink tokens and wallets, we can also reframe certain categories of NFTs and membership tokens. In particular, we could introduce a norm not to transfer NFTs and POAPs issued by reputable institutions that honor conference attendance, work experience, or educational credentials. The transfer of such membership tokens (if value traded) may reduce the reputation of the wallet and may prevent the issuer from further issuing membership or POAP tokens to the wallet. In a non-custodial ecosystem, a large number of users acquire considerable financial reputation and hold stake in their wallets, which can serve as indirect collateral that they do not abuse the expectation of non-transferability.
While all of these pathways face their own challenges, we hope that, through methodologically diverse sets of small steps, we will increase our chances of converging to a quasi-equilibrium state in the medium term.
first level title
