Today we announce the Reputation Network, an open source project on which we have been working for a good year. The reputation network aims to help people leverage their personal reputational data for things that matter to them, like getting credit or access to homes.
Understanding the Reputation Network
In the past, trusting was simple. We only interacted with a limited amount of people who had an expectation of our behaviors based on information symmetry, homophily, common relationships, and past transactions. As cities grew, we could not know everyone, and so we started to use proxies, like brands (some people will prefer McDonalds when they are abroad rather than a local restaurant they don't know).
Your reputation is an asset, but it is difficult to prove and port when you move into new circles. This leads to recommendation networks, which often have mirror biases (hence the whole white male Stanford grad VC), Matthew Effects (rich gets richer), and other biases.
We propose a network where people can more easily and certifiably prove their reputations to other people or companies, to help them access the services or products they need. We do this by proposing an ontology that companies can use to share part of their data (the reputational data) about a consumer (for example, a utilities company would not need to share your address or how many KWH you use, but that you pay your bills timely 99% of the time), for the benefit of that consumer, when the consumer needs it.
On top of that, a marketplace of scores can make use of those data points to make smart considerations about people, taking hundreds of data points from different companies who share (and building some machine learning around it), or just simplifying one or two data points into something useful (e.g. a score that enables proving that your salary is higher than $2000 per month, but without showing your actual salary, which would be useful to get access to a retail credit whilst maintaining privacy).
Users benefit in two principal ways. Firstly, for being able to prove that they are trustworthy quickly, as their different actions contribute towards their profiles and scores. Secondly, they benefit from the filter proving reputation whilst maintaining privacy, as explained in the example above.
Companies benefit from contributing to their customer's reputational profiles, with two implications. First, those consumers have more incentives for good behavior than before (e.g. short term credit companies can expect a lower default rate). Second, those consumers will prefer to buy the company's services precisely because they now their actions will count and contribute towards their profiles, while using a competitor who is not part of the network would not help them.
Scoring companies are new players who can build new innovation on top of consumer data and charge clients for the calculations.
Technical Advantage: Personal Data Sharing done Right
Legal proofs for data sharing: Users have full control over their data, while companies have cryptographic proof that they have consumers’ permission to use their data.
Data leaks prevention. Unlike today’s Internet, data in the Reputation Network are shared with a purpose. Apps receive personal data to produce a particular output. Verifiers mediate in data transfers to make sure exchanged data conform to a pre-agreed schema without being able to read the raw, unencrypted data.
Keep private data private: Obfuscated identity prevents personal data from being identifiable by intermediate apps. This makes possible to build zero-knowledge reputation designs, whereby a reputation score can be computed out of different source apps, in a way that neither the final score recipient knows the source data, nor the score app knows who the data owner is.
Clear data accountability. You’ve already registered in multiple web sites and mobile apps that you trust, where you generate data which in many cases is not feasible to decentralize, or there’s little incentive to do so. The Reputation Network takes a federated approach by making these companies accountable for data storage and executing the data rights of their users.
Differences with Traity
While Traity has a similar mission, Traity as a private startup has some weaknesses that the Reputation Network can solve:
- The Reputation Network is federated. This enables any company to share their data when the user requests it, but keep it in their servers otherwise. This reduces single points of failure and should give corporates confidence in keeping their own data and IP.
- Reputation is contextual. A single score is never going to be enough. No single company or team can fully understand reputation. You can have a great reputation as a consultant but not as a cook. There are some principal components around honesty, but the context is so relevant that creating a marketplace was the best answer we could give, so that people can create increasingly valuable scores for different demographics and use cases. You could have a yield score for lithuanian farmers, a validation of number of hours dedicated as a red cross volunteers, the treatment quality of physiotherapists, or a credit score for drivers that confirms they have done more than 100 trips last week on Uber.
The most important next steps have to do with working with corporations to participate in the Reputation Network and make some of their data available. The larger companies will probably prefer to start with our testnet and test data, but smaller companies can jump directly into the network and start to use it and benefit from better accountability loops with their customers.
If you are interested in what we are doing and would like to participate, feel free to take a look at the SDK or papers, or reach out to us and we will be happy to help.
When moon? Will the value of the token go up in price?
No. There is no token. Payments run on ether. This is an SDK.
The reputation network is identity agnostic, and we are trying to make it platform agnostic, so that we can turn from Ethereum to Hyperledger or other platforms and enable different companies whose CIOs have different compliance challenges.