Rethinking TCRs: KILT’s concept of the Token-Curated Attester (TCA)
Token-Curated Registries (TCRs) were conceptualised as decentrally managed lists with intrinsic economic incentives that, through the wisdom of the crowd (bottom-up), provide truthful information that cannot directly be verified. We at KILT would for instance search for an answer to the question “Which is the best restaurant for business lunch in Berlin-Schöneberg?”, as the offices of KILT Protocol are located in this district of Germany’s capital.
In this article we argue why we believe that a TCR is not the appropriate technology to provide us with a reliable answer to questions of this kind. We introduce a concept which we believe has a realistic chance to solve the issues of the current bottom-up trust systems, improving on the original Token-Curated Registry idea: the Token-Curated Attester (TCA).
The TCR Design Flaws
But let’s first have a closer look at the properties of a TCR. Aleksandr Bulkin’s article about “Token-Curated Registries that don’t work” elaborates in-depth on the shortcomings of TCRs as they are described in the TCR White Paper. He identifies three criteria that, due to human nature, must be fulfilled in order for a TCR to be successful:
- There is an objective answer to the particular question.
- The answer is publicly observable.
- The answer is very cheap to observe.
Now let’s apply the criteria one-by-one to our “Best restaurants for business lunch in Berlin-Schöneberg” list. In this example for a TCR, curators (token holders that govern the list) gather to vote on restaurants which applied to be on the list by stating their names and paying a deposit.
- The answer whether a particular restaurant shall belong on the list is not objective but clearly subjective, e.g. due to individual taste or notions of “good service”. (-)
- The information is available to the public, as guests can visit the restaurant during opening hours. (+)
- The information is not cheap to obtain. Voters need to go to the restaurant and try it out at least once, which usually goes along with time, effort, and a bill to be paid. (-)
That’s already two out of three of Bulkin’s criteria not fulfilled. Furthermore, we know that in a TCR, investors curate the candidates and act as experts. In the voting process, it is the highest investment that has the biggest chance to win. But investors are not necessarily good restaurant testers, as potentially anybody can take the role of a curator by investing into the TCR and there’s no quality assurance. Consequently, that list might actually not serve its original purpose of providing truthful answers but rather mirror subjective opinions.
Best Practice from the Analogue World
When it comes to solving complex subjective problems, in the analogue world we observe a different, economy-driven mechanism. In order to make decisions, i.e. to answer questions, organisations draw on experts. Good decisions have a positive effect on the organisation and thus enable it to compensate the experts for their work. The added value of the actual organisation is the curation of a list of competent experts who can maintain or even leverage the positive effect on the organisation by making the right decisions.
In this approach, the curators are an organisation, a company or a brand that stands for a certain expertise, perspective, and trustworthiness. They increase the value of their organisation with their employees and freelancers (e.g. by curating a list of experts). Those experts are attracted by incentives, e.g. money or reputation, provided by the curator organisation. The more successful the curator organisation is, the greater incentives it can offer.
A restaurant tester (expert) would choose to work for a high-profile restaurant guide (curator organisation), because it is able to pay her well, compensate for her expenses, and because its prestige is beneficial for her personal reputation. The expert would also be careful in her decisions since doing a bad job results in being dismissed (expelled from the list by the curators), damaging her reputation, and finally losing her income. Therefore, this system disincentivises bribed or bad decisions.
A restaurant (claimer) claiming that it’s the best spot for business lunch in Berlin-Schöneberg would apply to the organisation. The experts would then execute their complex job of testing the restaurant on behalf of the organisation. They would deliver the result of their work to the curators, who would publish it under their company/brand name on the list. Good decisions by good experts increase the value of the recommendations for those relying on them as well as the value of the curator organisation.
As we see, the basic principle of TCRs which allows to make decisions through a curation process is to be seen very positive and useful if we decouple the curation from the actual work of making complex decisions. This means that the decision whether to include a restaurant on the list should be done by experts in the relevant field and not the curators. The curators, on the other hand, need to enjoy trust in the relevant community, because the experts’ work will be published under their name. This trust is essential for all involved parties — only if the list created enjoys trust in the relevant community, people will use the recommendations it contains.
KILT Trust Structures
At KILT we deal with trust structures on the internet. One possible definition of trust we find very to-the-point has been made by Alexandr Bulkin:
“[T]rust is defined as a state where subjective information can be reliably assumed to be objective. Bob tells you that Alice is married. If you trust Bob, you will believe him and will not ask Alice out. If you don’t trust Bob, then you might decide that Bob simply wants to eliminate a competitor.” Source
KILT proposes a concept of trusted entities (“Attesters”) with Claimers being able to state attributes about themselves, e.g. “I am a fair-trade chocolate” or “I offer the best business lunch in Berlin-Schöneberg.” Attesters can confirm them and issue the corresponding Credentials. The Credential remains in the possession of the Claimer who can show it to any “Verifier” upon request. If the Verifier trusts the Attester, he can also trust the corresponding Credential presented by the Claimer. Credentials are stored as hash values on the KILT Blockchain by the Attester. They can be revoked at any time if the Attester finds that the particular attribute does not apply any longer to the Claimer.
→ For a more detailed process description watch our KILT explanatory video here.
In many use cases, the Attester already enjoys public trust. Attesters can be trusted individuals with a high reputation amongst Verifiers or trusted organisations, such as companies, governments etc. At KILT we call this “Top-Down Trust” and it’s the basic concept that we built our protocol on.
→ If you’d like to get an idea of how it works, you’ll find our KILT Demo Client here.
However, there are use cases where the trust of the Verifiers in the Attester has to be built up first. A good example is a newspaper. After it was issued for the first time, it has to build up trust with its readers over time. Through good journalistic work like being fast, entertaining, telling the truth, looking at different sources, etc. the (brand of) the newspaper will eventually gain the trust of its readers (Verifiers). In the KILT vocabulary we call this “Bottom-Up Trust”.
In this newspaper example it is not the publishing house (Curator organisation) that writes the articles. It just maintains the brand, cares for the commercial aspects and, most importantly, curates a list of excellent journalists (Experts). These Experts do the actual writing. Doing their work as bloggers and publishing it under their own name only, they would be less likely to make a living from it. But by doing “good work” and publishing it under the newspaper’s name they increase the value of the whole brand and will get rewarded for this. That’s why journalists (Experts) are highly incentivised to work for a high-profile publishing house (Curator organisation).
Token-Curated Attester (TCA)
Until this point we have evaluated the proclaimed weaknesses of the TCRs. We’ve reached the conclusion that the original motivation to design it was actually a very good one: providing truthful information. We then took a look into how similar problems are solved in the analogue world and figured out that society has already developed a very smart solution. Now let’s combine these findings and see if we can apply them to KILT Protocol.
- take the technological principles from TCRs,
- decouple the curation from the decision-making part as in the analogue world,
- and then implement the resulting technology into KILT Protocol, replacing the Top-Down Attester,
we receive a very powerful concept that we named “Token-Curated Attester” (TCA). The TCA comprises the two different roles “Curator” and “Expert”.
We currently conceptualise the TCA just like the Attester in our Top-Down approach:
The Claimer can send his attributes and ask the Attester for legitimation and the cost of the Attestation. The TCA automatically replies with the price. If the Claimer accepts, he sends a (Claimer-) signed version of the Claim, accompanied by the necessary funds.
At this stage, one or more Experts (depending on the configuration of the respective TCA) pick up the task and take all necessary measures to check if the Claim is legitimated to get certified. If this is the case, the Credential gets signed and sent to the Claimer. A hash value of the Credential is stored on the KILT Blockchain on behalf of the TCA. If the Expert decides that the Claimer failed to meet the criteria to receive the Credential, the process ends without a Credential issued. The fee paid by the Claimer remains in the TCA and is not returned in any case, as it covers the costs of the legitimation check.
At KILT we are still working out the technical concept for the Token-Curated Attester and we think it could be used for numerous business models. TCAs could have the potential to enable virtual structures, such as companies or departments of companies, running on a blockchain. Those virtual structures could be completely transparent for investors and for those working for them, while being regulated and taxed by governments like traditional companies. That’s why we also intend to build a TCA Factory on top of KILT Protocol, which would allow anyone to set up a TCA on the KILT Blockchain by clicking through simple customising options.
We will soon provide you with more information on our ideas and our progress. Until then you might want to check out our upcoming new version of the KILT White Paper for further details. We happily receive your feedback and invite you to share your thoughts about TCAs with us!