Scalar Capital partners discuss decentralized identity and reputation

Original Title: "Decentralized identity and reputation"

Original Author: Linda Xie

Original Translation: BlockBeats-HQ

Linda Xie, co-founder and managing director of the blockchain investment firm Scalar Capital, and an early product manager at Coinbase. Scalar Capital, a digital asset investment management company, was co-founded by Jordan Clifford and Linda Xie in 2017, headquartered in San Francisco, USA, with investments in projects such as DFINITY, 0x, NEAR Protocol, StarkWare, and more.

One of the most exciting topics in the crypto space is identity and reputation systems. People can verify their identity and reputation and apply them in decentralized applications (dapps). This is a complex topic, so this article addresses some of the challenges faced by decentralized identity and reputation systems, as well as ideas for useful crypto applications.

Identity

I have communicated with many identity system teams dedicated to developing compliance and anti-fraud tools at Coinbase. In fact, accurate identity verification is quite difficult, especially in countries with little infrastructure to support it. There are many different regions and types of IDs, leading to a large number of IDs. It is hard to determine whether someone providing unconventional IDs has presented legitimate identification, which is why identity verification companies often require users to provide a range of documents that are accepted within a limited scope.

IDs are easily stolen and can be easily purchased on dark web markets, so you cannot always know if the person providing the ID is a real person. Therefore, some service applications ask a series of questions, such as which street you grew up on, what addresses are your actual residences, etc. But this also brings other issues, such as some people not having enough history to answer the questions, some forgetting the answers, or this information being easily discovered and abused by those trying to impersonate them.

Moreover, there are 1 billion people without official IDs, and half of the women in low-income countries lack IDs, which excludes many from services that require official verification.

Residents of low-income countries (LICs), especially women, are significantly affected by the lack of IDs. Source: World Bank

To address these issues, it is necessary for governments to have better systems for issuing and tracking official IDs. One area where crypto technology can play a role is ensuring that the person using the application is the correct identity. When a person signs a transaction using their private key, decentralized identity can prove that the operator is indeed themselves. Note that this also means people must protect their private keys or need a centralized company to abstract this process.

Zero-knowledge proofs can also be used to verify identity while ensuring privacy and protecting sensitive information. For example, if a person provides an ID to an application, that application will check whether the ID is from a sanctioned country without providing more details about the ID. Another interesting thing you can use to prove is that when you have a certain net worth to become a qualified investor, you do not need to provide details about the total amount of assets you hold, such as cryptocurrency account balances.

Reputation

In the U.S., there are currently multiple credit bureaus (such as Experian, TransUnion, Equifax, etc.), and other institutions rely on these bureaus to obtain personal credit information. Credit bureaus can disadvantage certain groups, such as international individuals and young people. Peer-to-peer lending companies like Lending Club and crypto companies like BTCJam have attempted to improve these systems by adding additional data points. Reputation systems are very useful in the decentralized finance (DeFi) space. Currently, when someone needs to borrow from DeFi, they need to over-collateralize the loan so that if the borrower defaults, the lender has enough collateral. Reputation systems can help reduce collateral requirements. Most activities in DeFi are public, so you can build a credit history by tracking loan amounts and repayment status. In the future, companies may emerge that aggregate this data and provide credit ratings, making credit bureaus not the only entities that can control this information.

However, people can spend a lot of time building their reputation to later sell or scam larger amounts of funds, so it is important to have appropriate controls, such as collateral requirements and restrictions. People can also create multiple accounts and build reputations for all accounts (e.g., by interacting between accounts), so it is necessary to check whether these accounts are controlled by the same person. One way to check is by requiring someone to connect their public key repositories (Keybase), GitHub, Twitter, etc.

Cryptocurrency exchanges with fiat channels face a particular fraud issue, as crypto transactions are irreversible, but the payment methods for purchasing that cryptocurrency are reversible. Providing fiat channels without a strong anti-fraud system may risk the entire business. Therefore, data from these crypto exchanges can identify which accounts have a good activity history and which accounts may be fraudulent. Some of these companies may provide fraud/reputation scores (which users will ultimately choose themselves) to help customers provide data points to other services proving they are not fraudulent users.

Websites like hive.one and fifty.one can also serve as excellent initial resources for developing reputation systems. These sites provide credibility scores for accounts based on the influence of their followers.

Screenshot of hive.one website

(Note: hive.one is an algorithmic platform developed by a group of crypto-conscious researchers, engineers, designers, etc., providing relevance scores for users' social channels; it starts with establishing a baseline scoring mechanism called peoplerank from Twitter, working similarly to Google's pagerank, but the algorithm does not apply relevance scores to web pages but to conversations or messages sent through social channels; fifty.one is an algorithmic platform that categorizes, scores, and ranks individuals, projects, and news based on Twitter influence in the crypto community using Twitter data.)

These reputation scores can even serve as part of the token voting weight, so that those with a large number of tokens do not control the entire voting system. However, a balance needs to be found so that influential people do not become the ones who can make all decisions. There are many interesting experiments in this mechanism design.

You can also establish a reputation system that allows people to vouch for others and develop a reputation network based on this system. If the behavior of the person being vouched for is negative, it will also negatively impact the public reputation of those who vouched for them. However, this is a very complex model, and to encourage people to risk their reputation, the right incentive mechanisms must be in place.

Another way is for people to bet on a company's or individual's reputation score in prediction markets. If something goes wrong, it can be used as a hedging mechanism. However, this is more like a dystopian thought experiment, as it introduces some improper incentives. For example, one could make money by spending time damaging someone's reputation.

Conclusion

Overall, developing high-quality identity and reputation services is a challenging issue. I believe the ultimate outcome will likely be a combination of centralized and decentralized systems. Without a centralized entity to ensure the accuracy of this data, linking real-world identities to the crypto network is particularly difficult. For example, you could use a centralized identity service and then integrate it into a crypto application. You could also collect information through some centralized companies, but no single company can control people's reputations or own the documents. I am excited to see more experimental approaches in this area, and I am very willing to engage with people working in this field.

Thanks to Jordan Clifford for reviewing this article, and to Ali Yahya for sharing some thoughts on this article.