The Encyclopaedia Brittanica is the longest running in-print encyclopaedia. It was printed for over 244 years, up until 2010. Its discontinuation can be attributed to the rise of the internet. Today, with Wikipedia, anyone with an internet connection can read about and contribute to any topic with the click of a button.
The internet created the first iteration of a “World brain” (H.G. Wells), a global body of knowledge. It greatly broadened access to information, and gave us all a voice. Today in crypto we see experimentation around communities, decentralised autonomous organisations (DAOs) and prediction markets. Web 3.0 is taking human coordination and collective intelligence to the next level.
Kleros is a crypto protocol which is leveraging the crowd to resolve disputes. By combining game theory (mechanism design) and crypto-economics, Kleros is proposing a radical new way of resolving disputes. In the process, it hopes to redefine our justice system by bringing justice to all.
Why open online dispute resolution?
Dispute resolution is key to exchange and trade. It is the process of resolving disputes and conflicts between parties. Contracts are incomplete. And yes, even smart contracts! We cannot foresee all eventualities when signing a contract. Therefore at time of enforcement if a contract is unclear or broken, we rely on the justice system for dispute resolution. If a dispute arises it can be ‘resolved’ in different ways. Litigation is the most common way, which is based on the judicial system and applies the law in the name of a state.
We are increasingly transacting and operating online. Several trends highlight the inadequacy of the traditional legal system in the digital age:
- We are operating at a global stage. However, the traditional justice system mostly remains local. If I hire a 3D artist from a country at the other side of the world, what court do I go to, to resolve a dispute?
- We see a rise in small value transactions. Traditional courts are often too costly and inefficient to cover low value claims.
- Disputes are increasingly complex, focused on niche topics. The rapid advance and diffusion of technology is creating many new fields which require specialised knowledge and expertise. What court is apt to make an informed decision about 3D art and tooling? In crypto we are witnessing this first hand. Disputes around smart contracts, NFTs etc all require unique and constantly evolving knowledge and expertise.
Web 2.0 to 3.0
In Web 2.0, i.e. e-commerce marketplaces, faced this same issue. The old judicial justice system was too slow, expensive or lacked the flexibility needed to cover disputes between users. Therefore, they relied on alternative dispute resolution systems and created their own internal systems to resolve disputes. This enabled them to handle disputes faster and at lower cost. And this was key to their success. Reducing ‘transaction costs’ made their business models viable to begin with.
Web 2.0 platforms created an internal safe space (walled garden), where transaction can take place. However, these platforms set the rules (body of law) and can always overrule any decision. Furthermore, on these platforms users are locked-in. They lose their reputation, identity and assets if they decide to switch to a new platform.
With Web 3.0, we are moving from walled gardens, where the user is the product, to open protocols where the user is a co-owner. In this new paradigm, a user regains sovereignty over his/her data and assets. And he/she is able to fluidly move between all platforms in the stack with no penalty. In Web 3.0, terms of service, rules and contracts still exist. However, these are created and handled in an open and distributed manner. Kleros wants to build an open dispute resolution system which is built for and native to Web 3.0.
How Kleros works
Kleros connects parties in a dispute to a court of crowdsourced jurors. Crowdsourcing jurors is not new and was already used by some Web 2.0 platforms. On Taobao, Alibaba’s C2C marketplace, volunteering users handle most of the disputes. However, Kleros is built on Ethereum (public blockchain) and is thus permissionless. This means that anyone: parties seeking dispute resolution services but also jurors can use it without permission. And they do not run the risk of being deplatformed. Furthermore, Kleros or any other single agent cannot overrule the outcome of a ruling. Finally, the escrowing of funds and enforcement of incentives are handled by smart contracts. No third party needs to be trusted for this.
Parties in a contract
Two parties entering into an agreement, need to agree to use Kleros for dispute resolution when setting up a contract. This also entails that funds will be kept in escrow. Such a clause can be added to a smart contract and can be abstracted from the user through an interface. In the case of a dispute, the funds are released based on the ruling. This ensures enforcement after arbitration.
In the traditional justice system, jurors and expert witnesses are both selected and vetted. With Kleros, anyone can become a juror by staking capital ($PNK, Kleros’s token) in a court. Here, jurors wear a double hat. They are both experts and decision makers (jurors). Kleros has different courts, which cover different types of disputes and require different expertise. These courts have unique staking requirements, rewards and may have subcourts. For example: someone can stake in the general Blockchain court, but also in one of its subcourts: ‘Blockchain non-technical’ or ‘Blockchain technical’.
Jurors in a court are selected at random for each case. After being presented with the case and the evidence, the jurors get to rule on the case by voting. The majority determines the outcome of the case. Jurors which do not vote coherently with the majority lose their staked capital. Jurors that vote with the majority are rewarded. Jurors are thus incentivised to only participate (stake capital) in courts in which they have the required expertise. This court voting system relies on the concept of Shelling points (Game theory). You can read more about how this works in the context of Kleros here.
After a dispute, if a party disagrees with the decision reached by the jury, he/she can appeal (at a cost). The case is then retried with a new set of jurors. This time there will be double +1 the amount of jurors. The case may go to a higher level court. For example an appealed ‘Non-technical Blockchain’ case, may go to the larger ‘Blockchain’ court.
By using randomness, the appeal mechanism and incentives, Kleros wants to offer an efficient solution to reach a fair ruling. This makes it difficult and (too) expensive to cheat the system (through bribes etc). See the Whitepaper for more on the mechanisms.
Kleros in practice
Kleros can help reduce the burden of the existing justice system and can give a place to a set of cases which are currently not handled by the old system (i.e online disputes). While Kleros offers increased efficiency for many types of disputes, it is not always the best option. High stake cases, where subjectivity and fundamental rights are at play, are better left to the traditional system (for example: criminal justice). As a Web 3.0 native protocol, it lends itself well to online disputes where there is an abundance of use cases.
A minimally extractive protocol
Kleros coordinates the exchange between suppliers (Jurors) and consumers (parties in a dispute) to resolve disputes in an open manner. Two parties entering into a P2P contract, can use Kleros’s escrow application. However, Kleros can also be integrated by other platforms. It can then be used in the background without being visible to end-users. Kleros wants to create a public infrastructure layer for dispute resolution, that any platform can interact with. As a protocol, it strives to be minimally extractive. This entails that it coordinates the exchange between suppliers and consumers of the service without rent-seeking. By using Kleros for dispute resolution, a platform can focus on what it does best and is not constrained to its own user base for disputes. Kleros offers access to a larger set of experts, that can organically grow based on demand. And it gives users (jurors) the ability to fully capture the value of their expertise. Ultimately, this can make the dispute resolution process much more efficient.
A collective decision making protocol
Disputes can take many shapes and sizes and can involve individuals but also entities. Jurors basically need to make decisions based on information and data. There are thus many potential use cases for a protocol like Kleros. The use cases can go beyond disputes in the traditional sense.
E-commerce and freelance marketplaces are obvious examples where Kleros could be integrated. Kleros created Linguo which is a P2P (freelance) translation application. Translators escrow money and are challenged by reviewers (escrow money as well) if they make mistakes. In the case of a dispute between the translator and the reviewer, the case is taken to the relevant Kleros translation court.
Insurance companies need to verify claims for reimbursements made by policyholders. They need to decide which claims go on the ‘Approved’ list, and which ones are declined. Here, misaligned incentives often lead to skewed decisions. Similarly, token listings on Coinbase/Binance need to follow a specific verification/approval process. Kleros curated lists allows for the creation of crowdsourced lists. This allow for the fact checking of claims without using reputation.
Kleros piloted this with the Kleros token list, which is a curated list of crypto-tokens. Only tokens which match a defined criteria (ex: not scams, specific track record etc) are allowed on the list. Users can add tokens to the list which match the defined criteria. Another user can challenge a previously added token if he does not agree. Similarly to Linguo, all parties stake capital and are thus incentivised to be truthful. If a dispute arises between two parties, the case is taken to the relevant Kleros court. This list is thus well suited for decentralised exchanges, where tokens can be listed without permission. This list is now used by exchanges such as DeversiFi and Uniswap (as one of the options).
Currently content and data on most sites is filtered (moderated) by a third party entity or its algorithm. For example Facebook or twitter may decide what content to hide, users to ban etc. Arguably, these platforms have too much power. A system like Kleros, could offer a more open and transparent way to moderate and fact check these platforms. Content that is flagged could be taken to jurors. Interestingly @Jack, is aware of Kleros.
A Web 3 lego block
We discussed how Kleros could greatly improve existing marketplaces and platforms. With Web 3.0 completely new organisations and dynamics are emerging. If successful, Kleros will most likely play an important role in this new paradigm.
As we mentioned earlier, smart contracts are also incomplete contracts. It is impossible to assess every possible situation that could arise when creating one.
Massive adoption of smart contracts requires an “escape hatch” mechanism for when strict compliance would produce undesirable or unfair consequences. But how to create a procedure for an escape hatch without using a centralized decision maker that introduces a new single point of failure into the system? (Frederico Ast, 2017)
A decentralised dispute resolution layer could help enable the widespread adoption of smart contracts. Furthermore, there are many smart contract use cases relating to the ‘physical world’. There needs to be a trustworthy way to bring off-chain data on chain. Here a collective decision making protocol like Kleros could fulfil the long tail of off-chain data needs. It could thus offer Oracle type services, as a complement to protocols like Chainlink.
DAO constitutional court
A potential use case envisioned by Kleros, is that of a supreme court for DAOs. A DAO could choose to use a system like Kleros if it wants to uphold a social contract. If members of a DAO make a decision through traditional governance that goes against “its constitution”, the case could be taken to Kleros. This could be a way for a protocol to ensure the dependability over the long term over some shared values or features, while still being able to rely on community governance for most topics.
In a similar vein, Kleros could be used for disputes that might arise between different DAOs. Many protocols and platforms are integrating and partnering with each other. It is very likely that as DAOs continue to take over traditional organisations, we will see way more inter-DAO activity.
While most integration we see today with YFI for example are pure smart contract integrations, some elements of a partnership are not only based on code. A DAO may enter into an agreement with another DAO based on certain conditions, shared values etc. Kleros may be a good arbitrator in a situation like this one.
Lists and Identity
Kleros is currently creating the Proof of Humanity list in partnership with Democracy Earth. The goal is to create a curated list of verified/real humans using Kleros. The list itself can then be used as a registry to distribute universal basic income for example.
In its court system, Kleros intentionally focuses on staking capital, instead of using ones reputation. With the proof of humanity list, Kleros is experimenting with new variables. The curation process is hardened. Users stake their own identity/spot on the list in order to verify/vouch for other humans. If it they are dishonest, they lose their spot and the benefits that might come with that.
On-chain data is the record of our of on-chain activity. For example one can transparently see what protocols an account interacted with. Applications like Rabbithole (On-chain Earn.com) and Alpha blue’s degen score are taking the use cases of on-chain data to the next level. In the future, Kleros may also want to leverage this data to further ‘harden’ the system and potentially reduce the need for upfront capital required by jurors. Some ways this could be applied:
- One could filter jurors, only accepting jurors with an on-chain history into a court. For example, having deployed a smart contract might be a requirement for the technical blockchain court.
- One could alter the voting weight of jurors, based on their experience in a field
- One could alter the voting weight of jurors based on their voting history on Kleros and thus increase their potential share of rewards. Upshot which is building a decentralised Q&A protocol is using such a model.
I feel like Kleros hasn't yet gotten the full attention it deserves. It is one of the older projects in the space (founded 2017). Similarly to DeFi's promise: to bank the unbanked, Kleros aims to "bring justice to the unjusticed".
At time of writing, Kleros has resolved over 575 disputes and has 574 staking jurors (Klerosboard). It has also paid over 325ETH to coherent jurors. This is still the beginning though, and I am excited about Kleros's future.
While not everyone is ready to trust a protocol of anonymous jurors just yet, this will be (or is) natural in Web 3.0. This is also where Kleros is finding initial product market fit, and where it will continue to grow and experiment. I believe that an open, decentralized dispute resolution protocol will be key to Web 3.0's success. At the moment, Kleros seems best positioned to become the 'justice' component of Web 3.0.