Known Neuron Proposal: D-QUORUM

Governance NNS Governance
, ,
1

D-QUORUM neuron ID = 4713806069430754115

We’re excited to share this decentralization initiative with you… D-QUORUM is a special NNS neuron in that it’s owned by the NNS itself. This can be confirmed using the NNS governance canister.

steps to verify (click to expand)
  1. Click list_neurons to see API details…

image

  1. Fill out the necessary fields, including the D-QUORUM neuron ID, then click ∞ Call…

image
image673×548 19.4 KB

  1. Observe the controller of the neuron… rrkah-fqaaa-aaaaa-aaaaq-cai

image
image572×145 18.1 KB

  1. Verifying that this matches the NNS governance canister principal

image

This is a symbolic gesture. There’s no type of proposal that allows the NNS as a whole to vote and modify this neuron. However, the symbol is meaningful. The idea of D-QUORUM is that it cannot be controlled by an individual, and instead it faithfully follows the preferences of the NNS as a whole.

The NNS can hold elections for preferable NNS reviewers. This occurred for the first time several months ago.

NNS election results (click to expand)

The initial followee configuration for D-QUORUM is to:

  • follow DFINITY for every topic (except the private ‘manage neuron’ topic - see ‘Founding D-QUORUM Committee’ below), and additionally to
  • follow the elected reviewers for the special topics described above (see ‘NNS election results’).

This followee configuration can be confirmed using the NNS governance canister (by inspecting the same output that contains the controller information - see ‘steps to verify’ above).

Founding D-QUORUM Committee

As described, no means has been implemented for the NNS as a whole to directly modify a neuron (even if the formal controller is the NNS itself). This means that there’s a need for a committee to observe the will of the NNS, and faithfully ensure that this will is actioned on the neuron (by changing the followees according to election results). This committee will be followees of the ‘manage neuron’ topic, which is a special private topic that only the neuron followees can see, allowing them to modify the neuron by consensus. At most the committee can have 15 members.

So far this committee has representation from DFINITY, other ecosystem devs, in addition to well-informed and passionate members of the community. The full committee is yet to be confirmed and formally registered on the neuron.

These D-QUORUM co-founders will be announced soon.

D-QUORUM Co-founder Commitment

I will serve the NNS as a co-founder of the D-QUORUM known neuron, until such a time that I wish to step down from my role, and/or such a time that the NNS wishes all members of the D-QUORUM founding committee to step down (leaving the NNS with exclusive control of the D-QUORUM neuron). Until then, I will faithfully respect the overall preferences indicated by the NNS. These preferences are indicated either by NNS-wide proposal-reviewer elections, or by a mechanism yet to be established and agreed by the NNS. :saluting_face:

As a D-QUORUM co-founder, clicking I will in the poll below represents my alignment with this statement. :slightly_smiling_face:

  • I will
0 voters

(do not use this poll if you’re not on the founding committee)

Why Follow D-QUORUM

If you know exactly who you’d like to follow on a specific topic, and exactly why (not just because they’re already the most popular neuron) - then you should follow that neuron directly, instead of D-QUORUM.

However if you’re like most people, and either don’t know who to follow, or you follow the most popular neuron simply because they’re already one of the most popular, then you should think twice. This is the problem that leads to undue centralisation. It’s a network effect, powered by a feedback loop, and it stifles effective decentralisation. If this is you, then you should consider following D-QUORUM instead, as it’s designed to address this problem.

Assuming this pattern is inevitable (it’s been studied a lot), D-QUORUM’s mission is to provide the most decentralised neuron possible, such that it can afford to be the most popular neuron on the IC while still positively contributing to decentralisation of the network.

D-QUORUM allows users to say ‘hey, I’ve not got time to go in-depth on who to follow and why. I hereby donate my voting power to the DAO (for as long as I decide to follow D-QUORUM) so that the active and diligent members can distribute that VP with careful intent, via public elections’.

D-QUORUM (decentralised quorum) is a play on the word decorum (the proper way for something to be) - highly decentralised, and practically impervious to abuse, corruption, or centralised bias.

I :heart: ∞

More details (click to expand)

Network Effects & Follower-based VP
Network Effects & Follower-based VP800×2000 206 KB

6 Likes
2

There are a lot of valuable ideas here, but there are also a lot of red flags. Having a group of smart, active people get together to form a known neuron that controls itself in an intentional way for a specific cause is an awesome idea. We need more of these. However, representing this neuron as an being owned by the NNS is false and misleading and I fear will set dangerous expectations.

This D-QUORUM neuron is not owned by the NNS even if the NNS governance canister ID is set as the controller. The NNS doesn’t own or control neurons and it never should. The choice made by an individual neuron on how to configure itself, what governance participation strategy it follows, and how it decides to vote cannot represent the will of the NNS.
These choices represent the best interests of the neuron owners only. When we all vote on proposals, the will of the NNS is the aggregate result of all votes cast on that proposal. The entire NNS governance system falls apart if changes are made in the future to the NNS governance system where an individual neuron is actually owned by the NNS and is executing the will of the NNS.

In the event that there is actually an effort to change the fundamental design of the NNS to allow the NNS to control neurons, then this needs to be a bigger discussion that should be led by DFINITY. I believe it changes the framework of known neurons and the role they play in the NNS. Individual neuron owners would no longer be incentivized to form their own opinions and form their own voting strategies. The would simply follow the THE OFFICIAL NNS NEURON that is owned by the NNS, potentially making ICP staking a security. Hence, I would like to see an official response from DFINITY on this idea of NNS controlled neurons and whether or not we will move in this direction in the future. In the meantime, I would appreciate knowing if DFINITY sees any problem with D-QUORUM representing itself as being owned by the NNS and representing the will of the NNS. It seems like an official statement needs to be made about these claims.

This is also a mischaracterization. The NNS did not hold elections for preferrable NNS reviewers. DFINITY submitted Motion proposals because they wanted the NNS to decide who should be the first known neurons to receive Grants for Voting Neurons. Each neuron wrote a summary of why they are qualified and DFINITY asked the NNS to evaluate the 2 most qualified candidates for each topic. This was a trial grant program meant to bootstrap new known neurons to perform the work of proposal reviews. It was presented as a temporary system that would be replace in 6 months with something else. DFINITY did not imply in any way that the recipients of these grants should be configured as Followees for any specific neuron. They are simply grant recipients. So while it might be a good idea to follow some of these neurons directly because you know that they are getting paid to perform proposal reviews, it’s a stretch to say that the NNS held elections to determine preferable NNS reviewers.

Having a decentralized neuron is a great idea. Trying to convince everyone that they should follow D-QUORUM because it is owned by the NNS and reflects the will of the NNS is not so great. The D-QUORUM known neuron idea works really well when it is more accurately described as a decentralized group of like minded individuals who are coming together to configure itself intentionally with people and organizations who are known to roll up their sleeves and objectively perform the work of proposal review.

It’s important to point out that the Followees of this D-QUORUM neuron are not actually performing the work themselves in their role as Followees for Neuron Management for D-QUORUM, but they are paying attention to who performs this work and configuring the neuron to follow them. That makes it worth considering as a neuron to follow if it is accepted by the NNS as a known neuron. There are two other known neurons in the NNS who are already doing the same and who are already also decentralized (without claiming to be owned by or representing the NNS) and those known neurons are synapse.vote and DFINITY. These known neurons are controlled by a group of people who care about decentralization and choose to follow known neurons who perform the work of proposal reviews.

One other feature of D-QUORUM that needs to be addressed is that the intent is for all Followees to be configured by way of NNS Motion proposals. There is some risk with this approach since everyone in the NNS will be forced to cast a vote on these proposals without choice. Every Motion proposal carries 20x proposal weight compared to all technical proposals, so it is a major hit to voting rewards to abstain. Hence, everyone will need to vote including people who don’t think they should be required to vote on how another neuron is configured. These proposals will essentially be spam because they are forced on the NNS and will not be about the NNS. They will be about the Q-QUORUM neuron. Some may pencil whip a vote. Some may vote with nefarious intent. When DFINITY votes, they will trigger over half the voting power that will be cast on the proposal, so the decision will ultimately be made by DFINITY. It’s just a bad idea to try to use the NNS as a mechanism to decide how the D-QUORUM Followees should configure each topic. Instead, they should decide for themselves and configure it according to their own consensus.

Anyway, I will be very interested in learning what others think. It would also be good to hear from DFINITY as well as some of the founding members of D-QUORUM and whether they agree with all the ideas presented here by @Lorimer.

1 Like
3

I think you make some great points Wenzel. It might be that affiliating D-QUORUM with the NNS may cause confusion, or mislead some to view it as functioning in some official NNS capacity. I see it more as a communications issue. I believe Lorimer is emphasizing the mechanical fact that “it can’t be controlled by an individual.”

My personal view is that DFinity seems to view democracy too idealistically. Software cannot be built and deployed by committee, and a finished product requires thousands of big and small decisions to be made quickly and competently. Democracy slows this process down, and compromise typically sacrifices an optimal solution for the sake of stakeholder buy-in.

When I look at ICP governance projects, I look for competent, knowledgable, solution seekers who look out for the long-term interests of the project. The process is secondary. Good people can make a bad process work, and bad people can fail while religiously following the perfect process.

It’s the quality of the decisions, and their timeliness that matter.

I think Lorimer and the members of D-QUORUM can make a positive contribution to ICP governance. And I think they have the wherewithal to adjust D-QUORUM when they inevitably encounter some of the issues you identified.

I think this is more of a case of people needing their own space to express their own vision.

2 Likes
4

I’m thrilled to see the launch of D-QUORUM and the vision it represents for the ICP ecosystem. It’s critical to have decentralized governance backed by members genuinely invested in protecting the network’s core value: true decentralization. With D-QUORUM, users who may not yet fully understand the role decentralization plays in safeguarding the ecosystem have a way to delegate their voting power to a trustworthy neuron, one that reflects the community’s collective will rather than a centralized authority.

I believe D-QUORUM has the potential to be a reliable choice for ICP’s long-term stability and decentralized integrity, giving voice to those of us who prioritize a fair and balanced network.

3 Likes
5

Thanks for your input @wpb.

The NNS is verifiably the controller of D-QUORUM. This means the NNS is capable of making any changes to the neuron (if a proposal implementation is introduced to action that change in the future). D-QUORUM is the only neuron for which this is possible. It’s a new concept and I understand that it’s unfamiliar. Please let me know if you have any questions.

Note that the founding D-QUORUM committee are required given that a suitable proposal implementation does not exist for controlling D-QUORUM. If D-QUORUM is ever responsible for a major portion of total VP (due to followers) there will be an option for the committee to step down so that the IC can become highly decentralised. This is what we all want, and lack of this is one of the current drawbacks of the IC today. The IC is young, and has an exciting future.

Please be aware that anyone can propose a motion proposal, and anyone can vote on that proposal. This does not mean that everyone must vote manually.

2 Likes
6

Ok, have read all comments attentively.

1st, thanks Lorimer for this initiative and Wenzel for already sharing important risks / issues that it should mitigate.

A summary:

  • risk of misleading in the narrative of NNS.
  • risk of voting / election quality to be less than optimal or spammy.

I think, maybe a small pivot is needed. :thinking:

What if the narrative becomes: D-Quorum is an impartial elections committee, whose results can be followed on a Known Neuron with the same name?

The process can be:

  • open, with transparent recordings and discussions. Also with pre-approved processes at the founding committee and NNS level.
  • automated (as much as possible), as the process becomes clear and stable, an automated way using a canister(s) should be implemented.
  • standardized, contributing and actively discussing on the ICRC standards that @skilesare is already developing.

The only thing I worry about, is that Dfinity may already have plans for how to implement elections on chain (for SNS or NNS roles). @lara can you confirm if an “elections system” is being worked by Dfinity, or if community experimentation is actually valuable here.

In short, I am all up for a known neuron that respects and follows elections held at the NNS level.

We just need to figure out, what those “elections” are. :thinking:

2 Likes
7

I am not sure if it exactly meets what you are looking for and I don’t know if it will do much for a neuron that has already been given to the NNS, but ICRC-75 was designed to specifically give SNSs the ability to exactly this.

An SNS can deploy an ICRC-75 canister and add the Generic Functions for Managing lists of accounts, principals, and/or values. So an SNS could easily vote to add and remove members to a list and then the members of this list can ‘do things’. For example, a canister can be set up that looks at the list and only lets people on that list execute certain functions. Or it can effect a multisig where X number of people from the list have to approve of the calling of another function.

Please see more on this thread: ICDevs. DAOs Project Final Grant Post

2 Likes
8

This is interesting, thanks @skilesare. Does this differ a lot to an SNS registering a generic nervous system function that executes the manage_neuron function? This already allows an SNS to manage the permissions that different principals have on a specific neuron (and therefore the actions that they can perform). @RMCS has done some great work making this a smooth process with Toolkit. I’d be interested to learn more about your research into DAOs. I think the IC needs this sort of stuff.

That’s a nice way of putting it. Indeed, the committee need to be impartial and simply respect the results of an NNS election :slightly_smiling_face:

:heart_on_fire: I’d like to see this too. Ideally the founding committee’s role will be temporary (we’ll see what the NNS wants), and I agree that openness and transparency is top priority. I think all that’s needed for this to be fully on-chain is a proposal type that can execute manage_neuron on a neuron that’s owned by the governance canister (D-QUORUM). This is already possible for SNSs (generic functions can execute manage_neuron on the governance canister for other SNSs, and DFINITY has proposed introducing a native nervous system function that should soon allow an SNS proposal to execute manage_neuron on its own governance canister).

:heart_on_fire: This is very well put! Thanks @krzysztofzelazko

1 Like
9

From what I’ve understood, the D-QUORUM neuron (DQn from now on) will be configured to follow some neurons that are somehow active in the IC community and specifically in the NNS DAO.
In order to configure the DQn to follow neurons, proposals will be made to configure its followees and here’s where the committee comes into play. The committee will review and vote for the “Manage Neuron” proposals
I guess these proposals will likely be made by @Lorimer initially.
These proposals will reflect the result of the “elections” held by DFINITY in order to decide who to reward for the Grants for Voting Neurons initiative. In other words, these proposals will be made to make DQn follow the neurons chosen during the “elections” held by DFINITY.
There’s anyway space for anyone else to hold “elections” and make the DQn change its followees.

A few things that I don’t understand:

  1. What will happen when these “elections” done by DFINITY won’t be held anymore?
  2. What neurons will the DQn follow for topics that are not included in the “elections”?
  3. Why does the DQn neuron need to be controlled by the NNS Governance canister and not by e.g. a blackhole canister?

I agree with @wpb that this is not clear enough. The “elections” that were held a few months ago to decide what neurons would have received the grants were just done by DFINITY to ask the community who would have deserved the grants.

2 Likes
10

Thanks @ilbert, that’s right, this is the idea :point_up_2:

I don’t think there’s yet clarity on how long the elections in their current form will continue. Anyone could submit a proposal at any time, for example, to ‘unelect’ a reviewer who they strongly feel is not meeting expectations. It’s then up to the NNS what they do with that proposal (adopt / reject). It’s up to DFINITY if they would respect the NNS’ wishes, but for sure the D-QUORUM committee would be duty-bound to remove that neuron as a D-QUORUM followee.

DFINITY Foundation (as the only sensible fallback).

Without this, D-QUORUM would always need to be controlled by a small committee (or a limited portion of the NNS). By making the formal controller the NNS, it allows for the committee to be removed in the future (if desired by the NNS) and for the outcome of elections to be actioned entirely on-chain (as the result of an NNS-wide election proposal).

That’s correct, and D-QUORUM provides a convinience for followers (knowing that they can follow D-QUORUM and that this will be equivalent to following the election results over time - which would take effort to do manually, which most followers are uninclined to expend without a clear incentive).

Out of interest, do you believe the election results aren’t aligned with what the NNS considered to be preferable reviewers (ones that are considered best placed for the task)?

11

It only differs in that through an SNS generic manage neuron the whole SNS would be voting on that. This tool allows you to abstract certain things away so that you can implement other governance structures. For example, you could give the ‘marketing team’ a neuron that they have control over and the DAO can vote on the members of that list.

It may be a small thing, but the psychological difference between voting for Alice to be on the marketing team vs having your dao call the following function can be the difference between engagement and no engagement.

"(
    record {
      subaccount = blob \"${NEURON_ID}\";
      command = opt variant {
        Follow = record {
          function_id = 0 : nat64;
          followees = vec {
            record {
              id = blob \"\\ca\\0c\\08\\39\\1b\\89\\77\\84\\56\\42\\e1\\49\\32\\a7\\63\\b5\\31\\ad\\2d\\0b\\90\\b2\\cb\\93\\64\\51\\40\\ac\\68\\5c\\74\\67\";
            }
          }
        }
      }
    }
  )"

It also comes with a nice transparency dapp where people can go and inspect and investigate the lists:

image
image877×706 23.7 KB

3 Likes
12

I don’t get exactly the point of your question but no, I don’t think the election results aren’t aligned with the will of the NNS DAO members that voted on the motion proposals made by DFINITY.

2 Likes
13

Hey @lara will you please offer clarification on what you actually said on this topic? @Lorimer has claimed multiple times and in multiple chats that you explicitly said that DFINITY will implement his request for a manage_sns_controlled_neuron method. I don’t believe you actually said that, but he has interpreted it that way and has moved on to expanding the concept to the NNS. His design philosophy regarding all the D-QUORUM neurons that he plans to create center around DFINITY implementing this kind of method. Your full quote is below and you were providing a conceptual response to his leading question. You also indicate clearly in that post that you recommend he pursue the canister controlled neuron version of the options.

This conversation between you and Alex occurred before anyone expressed any opposition to the idea of an SNS governance canister controlling a neuron in it’s own SNS. You final remark to date in that thread (quoted below) is confirmation that he should pursue the canister controlled neuron route and that it is not clear that Alex’s preferred method should be implemented at all.

I think Alex is going to continue to claim that you came up with the idea of manage_sns_contorlled_neuron method unless you offer clarification otherwise. I would actually like to know myself if he is right because it is predictable how it would change NNS and SNS governance. There is no sense in engaging in governance with independent known neurons if in the future the nervous system will directly control neurons within it’s own nervous system. I would rather not waste my time now if this is our future.

Clarification: In case it is not obvious from my comments, I fully agree with the governance framework that exists today that a nervous system can and should be able to control neuron(s) of another nervous system with the one exception that the NNS should not control an SNS neuron. My opposition here is related to a nervous system controlling a neuron within it’s own nervous system.

14

I believe this to be the crux of the problem with Alex’s vision for D-QUORUM.

15

Thanks @skilesare. To make sure I understand, in this example the entire DAO would be voting for the members of a list. Is that correct? So the difference comes down to how the proposal payload (and outcome) are rendered?

16

Thanks for asking for clarity @wpb. I would add that I’ve not claimed to know what will happen. I have simply stated that @lara proposed the idea of a native nervous system function for this purpose (as a solution to the restrictions placed on generic nervous system functions). This is true. Remember that the SNS framework is designed to provide individual SNSs with flexibility with regards to how they can organise themselves. A reason to make it impossible for any and every SNS to call the manage_neuron function on their own governance canister has not been provided. This would be an imposition to limit the freedoms of all SNSs, and should be backed up with clear justification about the threats that would be mitigated.

Out of interest, have your opinions changed much since your TAGGR post, advocating for public reviewer elections :slightly_smiling_face:

17

Taggr has a very different governance system than the NNS and SNS. That post was related to Taggr where liquid democracy and neuron/known neuron concepts do not exist.

The NNS already has known neurons that must be registered much like node operators are registered. As I understand it, the DFINITY roadmap contains plans to expand the known neuron concept to require known neurons to register for specific proposal topics. I agree with this plan and have no issue with the NNS approving these registrations. It would make sense to me that these registrations are done via a new proposal topic analogous in many ways to the independent Participant Management topic instead of the Register Known Neuron proposal type that rolls up under the Governance topic, but I’m sure DFINITY is thinking through the best way to do it.

These registrations are very different than your claim that the NNS is electing preferrable reviewers via Motion proposals. Registering a known neuron for a specific topic is nothing more than an acknowledgement that they have committed to that role. It says nothing about their performance. It will certainly will not automatically set those registered known neurons as Followees for any specific neuron. I believe the registration should include a neuron ID, neuron name, neuron description, and a list of topics that the neuron is committed to reviewing and once registered the NNS dApp should reveal who has committed to which topic in the drop down for Add Followees when users configure their neurons. The default list should be those neurons that have registered for the specific topic, not every known neuron in the NNS. Regardless of how known neurons are registered, every individual neuron owner should be required to decide for themselves who to set as a Followee for their own neurons. There should not be a NNS proposal topic requiring everyone in the NNS to vote that sets it for them, including D-QUORUM.

In the ideal world, the NNS would incentivize many people and organizations to step up and register for every proposal topic. If the right incentives are provided, they will come. If your vision for D-QUORUM were to have the Followees for Neuron Management intentionally select the first 15 of these registered known neurons for every proposal topic and to remove those that you believe are not doing a good job, then I would fully support your goals with D-QUORUM and I might even be interested in serving on that council. However, that Followee selection should not be conducted by NNS wide proposals that manage the D-QUORUM neuron. If you build a canister controlled neuron for D-QUORUM, then that would naturally force a subset of the NNS to choose to join the D-QUORUM community in order to cast those votes. That would be an awesome service to the NNS community and I would have no issue with that community growing as large as you want.

18

Thanks Wenzel. Are you familiar with why TAGGR is designed without a concept of liquid democracy?

The core principle of D-QUORUM is that it’s maximally decentralised (as much as allowed by existing and future governance mechanics). It’s not dissimilar to a country voting for delegates that represent an average preference for the country.

I’d be interested to see @skilesare’s take on this, given his research into DAOs.

19

Removing my initial knee jerk response…apologies…

I agree that I’d like to learn more about @skilesare take on this topic since I know he has well reasoned opinions on how the NNS and SNS governance should work.

1 Like
20

This wasn’t my intention. I’m sorry it was taken that way.

Thanks Wenzel :slightly_smiling_face:

1 Like