Global portable identities on the blockchain: a proposal

Identity and rating systems have made tremendous progress improving behavior and leading to consistent positive experiences. Ebay, Uber, and Airbnb have used ratings to give an incentive for good behavior and weed out bad users from their systems. This is a large step forward from systems with no ratings, such as Craigslist, which allows bad users to misbehave over and over with new victims. The problem with the rating systems of the above website are that they are not portable — they are contained solely within the private databases of for-profit companies. They do not want to share them, and if they did, it may be in a way that individuals do not wish (sharing with advertisers or third parties).

Problems with current lack of portable reputation:
1. Losses due to repeated fraud
2. Missed transaction opportunities between honest strangers
3. Wasted effort duplicating positive reputations across similar centralized sites and applications
4. Scoring systems within a site do not have competition, and therefore lack the pressure to improve over time
5. Users feel profiled and disempowered
6. Site-specific scores do not incorporate additional valid information
7. Many existing reputation systems are either trivial to manipulate, or require a link to a real world identity
8. A company discontinuing a line of business can cause your hard-earned reputation with it to disappear
9. Companies with control over reputations have very high levels of centralized power

Therefore, I propose a new identity system that will have the following:

Proposed properties:
1. Free to create
2. Open to anyone to create or verify
3. You can always create a new identity or multiple identities
4. Global, portable, and independent of any one company
5. Can be anonymous, pseudonymous, or linked to one’s real name
6. Allows ratings from different systems to be imported and combined
7. Allows for custom formulas for aggregating ratings across services, weighting ratings based on the reputation of the specific raters, and considering the history and timeliness of ratings
8. Can be used on third party systems when appropriate
9. Open source and built on top of the bitcoin blockchain
10. Can help people demonstrate a history of trustworthiness and help people find trustworthy strangers to deal with
11. No one is forced to use the system unless they choose to participate
12. Individuals choose the information they wish to aggregate
13. Usable both online and in face to face interactions

The bitcoin blockchain allows any individual to create a private key (with a corresponding public address) and sign messages which are permanently written to the blockchain (a global, public ledger of information). We need a standardized method of designating a public address as a specific identity (and whoever has the private key can authenticate messages as coming from that identity). Information about transactions and reputations vouching for other reputations can be hashed and stored on the blockchain. A public address designated as a reputation will start out with no credibility (since it costs nothing to create, and the supply is unlimited, merely creating the address does not yet add anything over not using an identity at all). To initialize the identity, we must request an endorsement from another identity which already has a solid reputation within this system. Alternately, we can have the new identity “burn” bitcoins to deliberately create a cost. Burning bitcoins means sending bitcoins to an address for which no one has the private key, taking the bitcoins out of circulation permanently. This has a real cost for the burner, and has no direct beneficiary (except for the bitcoin community in a diffuse sense due to the reduced supply). Since there is no direct beneficiary, the burning cannot be faked (by sending funds to a party who will return the funds). Anyone can see bitcoin transactions, and thus the burning of bitcoins by the identity can be trivially verified anytime. The number of bitcoins burned is at the discretion of the creator of the identity. It makes sense to burn an amount corresponding to a non-trivial amount of common currency to demonstrate that the account was not created as one of many by a malicious user.

This can be implemented in code that allows for the creation and verification of reputations, as well as calculation of scores and integration with other applications and websites.

Consequences:
1. A fair system where trustworthy people can prove they are trustworthy to strangers
2. Users who have developed a long history of good behavior on one system can refer to it when starting on another system (consider an Uber driver switching to Lyft — must they start at zero reviews?)
3. Users can develop their own formulas or select from published formulas to decide how to value positive and negative ratings across multiple services that are attached to an identity, and formulas that have a better track record of predicting future behavior will become more widely used. formulas could weight events differently based on recentness, as well as relevance to the type of transaction currently being considered, and the size and consistency of prior transactions
4. Users can always start with a blank slate, building a reputation from scratch by creating a new identity, burning bitcoins, and behaving well over a period of time
5. Malicious users will not be able to develop identities with a long trustworthy history as this would be too costly in time and money
6. systems that do not have a built in reputation system could now take advantage of this system (a user on craigslist could post a link to their reputation on a reputation blockchain explorer, and digitally sign their listing with the private key of the identity
7. The private key to the identity could be stored in many secure ways, including biometrics or on specialized hardware, giving the identity a much higher likelihood of authenticity as compared to other forms of authentication (a signature, providing a password, publishing from a certain IP address or email address, using a secret code)
8. There is no unaccountable centralized party that has control over reputations
9. The silos of information across multiple systems can be made accessible
Issues:
1. Sale of pseudonymous reputations
While I do not currently have a mechanism to prevent this, the cost of establishing a reputation and the relative ease of destroying the value suggests that the sale of reputations to malevolent parties would be far less common than honest transactions
2. Inflating reputation with fake accounts transacting with each other
This can be mitigated by analyzing the integration of the accounts into the system as a whole, and weighting the reputation scores based on the reputation scores and reputation age of the respective accounts giving the ratings.
3. Less profitable for the creator of a system of this type than a centralized system
While this suggests that development of this system may proceed more slowly, many projects have been developed that are structured more for the benefit of society and users than the direct compensation of the creators (for example: Craigslist, Linux, and others)

Like what you read? Give Daniel Ameli a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.