HIP-68 - A Compromise to Forking: Implementation Strategy for a Second Arbitrator

Proof of Humanity
1 
HIP: 68
Title: A Compromise to Forking: Implementation Strategy for a Second Arbitrator
Author: arkayana.eth
Status: Phase-1 & in need of further contribution
Created: 19/09/2022

Simple Summary

Establishing an implementation strategy and the requirements for a second arbitrator, in addition to Kleros.

Abstract

I propose establishing a formal set of requirements, then seeking out an appropriate arbitrator which could operate alongside Kleros, rather than remove Kleros and the important services they provide to PoH. The second arbitrator can then provide profile verification services and dispute resolution as Kleros does which can be selected at the registration stage by the registrant.

This HIP seeks to be a comprimising alternative to the hotly debated HIP-49 that sought to remove Kleros entirely from the PoH ecosystem, recieving both avid support and rejection, as well as resulting in tremendous polarization in the community.

Motivation

Primary Motivation:

PoH is currently in a state of disarray. This has caused significant division and confusion among members of the DAO, mission board, and contributors, as well as a plausible, yet theoretical knock-on effect to the price of $UBI. The level of disagreement and personal gripe between key members of the protocol continues to disrupt progress and disenfranchise those interested in building on the protocol over the past year. Additionally, disecting the points of view on the forum and telegram channel is overwhelmingly challenging, and sourcing any productive, comprising dialogue is equally difficult, resulting in circular argumentation.

By having two arbitrators, this satisfies Kleros’ desire to remain an active service provider to the protocol, whilst also providing the ‘opposing view’ with an alternative arbitrator to use, satisfying their desire for a diversified arbitration team.

Secondary Motivation:

A second arbitrator allows for healthy yet friendly ‘competition’ in the form of comparable statistics benchmarking their own performance against Kleros’. For example, comparing rejection rates, registration rates, profile security, etc. This may result in collaborative improvement, evidence based critique, and increased effectiveness of arbitrators in the PoH ecosystem as they learn from each other and continue to build their own respective networks.

Tertiary Motivation:

Increasing exposure of the PoH ecosystem and the PoH DAO to other protocols, DAOs, and companies, thus increasing diversity in the PoH protocol and bettering our reputation. This is extremely important if we are to ever capture the continued interest of major industry proponents such as Vitalik Buterin, Elon Musk, as well as creatives, developers, and eventually politicians, or social entrepeneurs, etc.

Implementation

For this to be respectfully and democratically implemented, we need to first establish a list of requirements, taking into consideration the opinions of active parties of influence within the PoH ecosystem (i.e, Kleros opinion, Mission Board, Democracy Earth, and human voices in the DAO).

An incomplete list of requirements and goals for a second arbitrator:

  1. To provide adequate, provable sybil resistance in line with PoH DAO & HIP requirements
  2. To have publicly accessible data in order to monitor the effectiveness and fairness of the arbitrator.
  3. To have a point of contact or representative within the governing body of the second arbitrator to communicate with when necessary.
  4. To have a secure escrow system (if used)
  5. To have democratic and humanitarian interests at the core of their beliefs.
  6. To actively develop their arbitration system, even after implementation.
  7. For their operation expenses to be clearly illustrated and adequetly funded.

Rationale

To avoid forking.

In my personal opinion this should NOT be our first, actioned response to heavily divided opinion, I propose to establish the requirements necessary to integrate a second arbitrator into the protocol, satisfying a number of needs on either end of the argument.

Addendum: Whilst there are many solid arguments for forking as a concept, frequently highlighted by the member Shin in the “Go Fork Yourself” article, I do not believe this needs to be our first resort. Much of the emotional, antagonistic behavior inherant in the bipartisan environment that arose over the past year appears to be the main driver for forking, rather than it being a mindful, diplomatic resolution.

For that reason, I cannot support that the fork would be done in good faith or with clear intent, at least at this current stage.

I endeavored to be as neutral as possible in this proposal, and those reading should be reminded this is a comprimise, so neither party will likely recieve full satisfaction. I can only aim to achieve a happy medium in line with my interest in diplomacy and mediation.

If this proposal is considered to be a viable course of action, I will continue to ammend the “Requirements” list until a consensus can be reached before concluding Phase 1, and certainly before any contact is attempted with external arbitrators so that we approach them with professional clarity.

In the far off future, I theorize that three arbitrators would be ideal in terms of governance and representation, although this is a personal opinion and out of the scope of this HIP.

2

This would make the system strictly weaker as an attacker can just try chose the weaker arbitrator.
Kleros has been providing dispute resolution services to the DAO at an extremely low cost (less than 1$ per submission in average) without having been broken.
Some people disagree with rules and blame the arbitrator. The arbitrator just applies the rules, it doesn’t make them.

The main point of contention have not been about the arbitrator but about compromises between security and cost.
I argue that for the registry to be valuable, it must be extremely secure.
Other people argue that the registry should have a rejection rate as low as possible.

I am not interested working on insecure systems (and I believe other projects are not interested integrating those either). So if there is a side which want to lower the rejection rate below what is secure, it’s better to fork to make sure people who want to use a Proof of Humanity with the original, or even better security guarantees, can use it.

This proposal is not a compromise at all as it proposes to lower security.

1 Like
3

I agree, and I made that clear in the proposal.

If the second arbitrator meets the requirements set out in an implementation strategy by the DAO, a HIP, and all interested parties, including Kleros and yourself, especially in terms of security, then it wouldn’t be reducing the security of the network, only adhering to the set out standard. Hence why my very first point of requirement:

Additionally this is an Implementation Strategy , NOT an implementation. No implementation takes place before the arbitrator requirements are formally established.

Kleros provides an excellent service to PoH, but should it have sole proprietary rights over arbitration until the end of time? I don’t think so, nor is it a democratic ideal.

This is hypothetical conjecture and not a basis for argument - You are rejecting the idea of another arbitrator being involved under the assumption Kleros has and will always have superior security and that all other arbitrators are insecure by comparison. “I don’t want to” is the mainstay of your argument here.

I have personally used systems that offer more than one method of ID verification, this isn’t unchartered teritory, nor unheard of. Again, if a second arbitrator meets your own security requirements, there is no conflict with your ideals of security. Therefor we only need those requirements to be established, which you are more than welcome to suggest.

I disagree completely. It has been a HUGE point of contention ever since HIP-49 was composed, was absolutely a major trigger point in polarization, and many of the arguments inherent in the HIP are still discussed today. It is not the only point of contention, but it is veritably a major one, I don’t believe you can genuinely dispute that.

Lastly, I also have to highlight that you do have a conflict of interest here as both a Mission Board member and a key stakeholder in Kleros. Concious and/or unconcious bias exists, and business accumen would tell you not to let another arbitrator ‘have a slice of the pie’, figuratively speaking. Kleros’ business strategy is no reason for PoH to not consider the possibility.

2 Likes
4

It always make the system less secure as attacker can choose which arbitrator to attack. Actually even just changing the arbitrator would be more secure than adding an additional choice.

So the proposal doesn’t give any info. And I’d suggest that the best way for POH to fail is to use its limited resources to try to compete with the team which created it and is currently funding the project.

There is no proprietary right in a service provider. The simple fact that you could propose to use another arbitrator shows that.
But obviously, if a version of PoH doesn’t use Kleros, the Kleros Cooperative will not be able to provide funding or services to it.

First, you would need to exhibit an arbitrator more secure. And second, even if there was one, that would still make the system less secure as the security of system where one can choose his arbitrator is the security of the weaker one.

This is not a conflict of interest but an alignment of interest. Proof Of Humanity is a protocol created by the Kleros Cooperative as part of ecosystem development. I have therefore more interest in its functioning properly as if it didn’t that would impact me both as a founder of Kleros (as it’s a Kleros ecosystem project) and Proof Of Humanity.

1 Like
5

Some good points! Thank you for the reply.

It creates an additional attack point - true. However it also creates a contingency in that one attacker can’t use the same method of attack on both arbitrators.

In Scenario A with only 1 arbitrator, an attack on a weakness threatens the entire system.

In Scenario B with 2 arbitrators, an attack on a weakness threatens only the portion of profiles of the effected arbitrator. (say 50%)

I believe there are positives outside freedom of choice to explore. Of course there are negatives, as you suggest.

Slight overexaggeration? I listed key areas of importance, and various rationalizations - it isn’t just a blank space up there. I do stand by the choice to keep this in Phase 1 discussion for an extended period. As this is a very touchy subject, I am happy to edit this from a HIP to purely a discussion? Not sure on the rules there as it would leave a HIP slot empty. (I can just use it up when I next write one if nobody else does)

The conflict of interest is still present when it comes to another arbitrator being involved, as it could threaten your own role, especially if Kleros was assessed as the weaker of the two, albeit unlikely.

Although I do believe your intentions are genuine, I only want you to be aware of the potential of the same unconcious bias that impacts us all, myself included :pray:

I agree there! Said second arbitrator should absolutely have their own capability to fund and manage themselves, but as you suggest, their service to PoH also. Kleros shouldn’t have responsibility there, certainly. This expands on number 7 on my list of requirements.