Committee Neurons

[lastmjs]

@wpb

I hesitate at the idea of a community established roadmap for the NNS or for DFINITY. All roadmap discussions have focused on the DFINITY roadmap in the past, but perhaps it’s time to let that go. DFINITY owns their own roadmap and should be in full control over it. To me, they are clearly acting in what they think is the best interest of the IC and should be free to pursue those efforts as they see fit. They care about what the community thinks and they put a lot of effort into seeking community feedback, but that doesn’t mean the community should feel entitled to dictate their roadmap.

Yes DFINITY should be free to do what it wishes, but the NNS needs a way to come to consensus on what will be allowed into the protocol.

[lastmjs]

@wpb

What if your committee idea were integrated into a new sister foundation that could be formed to focus on governance and protocol topics that are important to the community?

Could be awesome! But not sure that alleviates the need for the same structure at the protocol level. Definitely interested in an organization like this though.

[Omaxus]

Nope
Sounds to me like forget decentralisation and do somethink like super representative.

A staker has always the option to follow someone with more expertise. The expert can get more following neurons when he/she is trustworthy.

So I will definetly vote against it.

[Forreal]

The rewards should come from the followers. Even a modest percentage of the incentives (e.g., 0.03%) would be enormous in compensating the committees performing the obligations and creating a treasury to conduct any independent investigation or analysis on proposals. So far, I like the initial concept.

[Choiseongin]

I agree.

Most people don’t understand what they’re voting for.

Let’s admit that people are chasing money to join the community and be ignorant.

We need to find a way to control members like parrots.

[wpb]

Wouldn’t that consensus come from voting on proposals? We need more public neurons by organizations that can prove themselves capable of filling the technical role and convincing the governing body that they are worth following? If those neurons existed today, then it could be as simple as removing a specific proposal topic from the All Topics Except Governance catch all category and then giving people a tokenomic reason to intentionally select a Followee for that topic. This is what happened to partially decentralize the Governance topic and it was effective. There just hasn’t been a reason for more organizations to form public neurons. I really think the solution to decentralization is going to be related to providing incentives for people and organizations to get much more involved.

[dfisher]

Upon reflection, I am not opposed to the idea of a small amount of existing inflation redirected into a NNS treasury. After all, with DAOs created on the SNS, they will all have a community controlled treasury. So why shouldn’t the NNS have a treasury?

That said, I don’t love the idea of an elaborate governmental structure that is employed by the NNS itself. Rule by committee at the NNS-level itself feels a bit scary and prone to corruption.

Rather, I think it would make sense for private organizations to apply for funding from the NNS treasury. They would have to demonstrate they have in place the right people for the job and the right structures to ensure transparency and accountability. Such grants I think should go to non-profit foundations who can use the funds to pay its committee members to further the goals of the non-profit.

This could open up a whole new set of possibilities. DFINITY would be just one non-profit contributing to ICP development. Imagine if CycleDao incorporated as a non-profit and got funding from the NNS treasury. Maybe they would have not bailed on the project…If organizations fail to perform their duties, or become corrupt, they can be cut off from funding.

@lastmjs @bob11 thoughts? It is worth noting that the amount of existing inflation this would require would go down over time as the price of ICP goes up and the ecosystem develops…

[lastmjs]

I’m just not sure the simple liquid democracy we have will ever be enough to satisfy the demands placed on the community.

As an example, in my mind the most important material aspect of governance currently is ensuring that the replica code and system canisters are upgraded securely. To feel comfortable with security updates, I want to see multiple (probably 2-3) independent entities to review all changes and sign off on them. How can we do that with the neuron structure we have right now? I wonder if it is necessary to create neurons with different capabilities than they currently have.

[lastmjs]

I like these ideas, but I think they can be implemented in tandem with committee neurons or some other upgrade to the governance mechanism.

The NNS having its own treasury makes a lot of sense in my mind, then organizations like DFINITY can apply for funding, or the NNS DAO can give out its own grants to the community (in addition to DFINITY).

[wpb]

@lastmjs

I wonder if it is necessary to create neurons with different capabilities than they currently have.

Yes…in a way. I think we need organizations (or people) that can specialize in certain proposal topics. There can/should be many of them. They need to be incentivized to do it, but their presence would give people an option for who to follow. In fact, public neurons who don’t claim to specialize in that topic can also choose to follow them. This structure exists today. All we need are incentives for neurons to specialize.

If I know your neuron ID and 14 other neuron IDs for people I know and trust on a specialized proposal topic…and you all commit to fulfilling the obligation of always voting on that proposal type (due to attractive incentives), then you, me, or anyone else could form a named neuron that specializes in that proposal type and promote it as an option to the community. I don’t even need to be an expert on that topic as the neuron manager. I just need to know who is who and manage my neuron followees with people I know and trust on that proposal topic and I need to publicize it effectively. This could happen with dozens of specialists and dozens of public neurons.

[dfisher]

It’s probably worth separating out the proposals for committees and a NNS treasury.

Personally I think the primary function of the committee should be to oversee the spending of the NNS treasury. The amounts of money will get large and at a minimum we’ll need people dedicated to making sure it is not wasted.

[dfisher]

Why is a committee that’s paid by the NNS treasury to ensure the replica code is safe and secure better than a private organization that gets funding from the NNS treasury to make sure the replica code is safe and secure?

Just want to make sure I understand the nuance of what you’re proposing

[lastmjs]

What I’m proposing requires changes to neurons, so that committee neurons can adopt proposals when their members vote, without regard to the rest of the NNS.

I’m proposing that replica code cannot be updated without the committee’s approval (or perhaps with a very large majority vote of the NNS as a check on the committee). I think it might be hard to ensure this using the current neuron system with outside funding.

The core component here I think is the committee neurons’ ability to unilaterally adopt or reject proposals.

[wpb]

@lastmjs

this proposal seeks to address those issues by harnessing the decision-making efficiency and effectiveness of smaller groups of expert leaders

Eva stopped by the ICP Maximalist Network telegram group yesterday and posted about a couple of new initiatives she and the growth team are going to start brainstorming. One of them was about Neuron Following in which she indicated she wants to look at ways to “motivate people to follow other neurons (experts) and conversely reward experts for following them.” There are similarities with what you have in mind regarding motivating expert participation in governance. You should stop by and search for her name (@ evafromgeneva on TG) and then read the commentary afterward. It sounds pretty interesting.

ICP Maximalist Network

[justmythoughts]

@lastmjs

These c-neurons would be tasked with reviewing the code for each update to each system canister. They would provide a check on DFINITY or any other party attempting to push changes to the system canisters not in line with the roadmap, or malicious in nature.

A good c-neuron to start with (as a test of this feature) would be System Canister Management, which is used to update the NNS canister.

@diegop answered many of my questions with respect to how DFINITY currently controls their neuron in this comment.

He also detailed a bit regarding the pager-type notifications each of the 9-voting members have for voting on urgent replica updates:

Assessing governance & voting enhancements

Quite literally, yes. They are all paged and they vote even at very odd hours. I am glad I am not one of them.

Putting together c-neurons that are gatekeepers to the actual protocol requires a much higher level of commitment than the current named neurons.

I mean no offense by this to any of the of the current named neurons, but being paged several times from 1:30am to 4am to get up to speed on an issue and make a coherent vote on it is not easy. It requires technical expertise on top of dedication. If a replica update gets pushed out and needs to be rolled back or a emergency patch fix, the c-neuron members need to be reachable and ready to go, even to the point of staggering vacation/time-off if membership is small enough.

As much as I’ve pushed back on this publicly, I honestly don’t trust the NNS community with voting at the code-change level right now. We’d end up with members who are whales and/or non-technical members of the community instead of those with specific expertise or at least high-level knowledge of the specific changes being made.

Realistically, I don’t think that DFINITY would approve of this either in the current state - even if you rally enough support in a governance proposal to approve of these changes, DFINITY would not implement them, and then any subsequent development efforts would be wasted since the community would not be able to gather enough voting power to pass a replica code change via the System Canister Mangement topic (where everyone currently follows DFINITY).

It’s for that reason that I recommend an the following experiment as a “half-measure”:

That DFINITY temporarily allow just a single, technically well-versed member from the community to become a voting member of the DFINITY neuron, to see if this community member can realistically assume non-governance topic voting responsibilities and meaningfully contribute to the DFINITY neuron.

DFINITY would allow the voting member in for 1-3 months to vote on a pre-defined subset of topics, and can obviously kick them out at any time DFINITY wishes to do so. If the voting member’s contributions impress DFINITY, then DFINITY may slowly gain more trust in the other members of the community to assume various responsibilities.

By DFINITY doing this, it would go a long way towards obliging and quieting the community’s concerns about not having a voice in the protocol, which has been a recurring and persistent discussion topic over the past few months.

I would nominate you, @lastmjs to assume this responsibility as a voting member of the DFINITY neuron.

[justmythoughts]

@wpb

Eva stopped by the ICP Maximalist Network telegram group yesterday and posted about a couple of new initiatives she and the growth team are going to start brainstorming. One of them was about Neuron Following in which she indicated she wants to look at ways to “motivate people to follow other neurons (experts) and conversely reward experts for following them.”

I’m not a big fan of rewarding neurons for how large their follower count is.

That sounds less like governance and more like social media.

[lastmjs]

I agree that we should scope this down to an MVP. I’m not sure starting with something as security-sensitive as the replica code is a good idea though.

Perhaps we should start with a roadmap c-neuron?

[wpb]

What makes you think rewards would be based on follower count? I don’t think that was said.

[lastmjs]

@justmythoughts

As much as I’ve pushed back on this publicly, I honestly don’t trust the NNS community with voting at the code-change level right now. We’d end up with members who are whales and/or non-technical members of the community instead of those with specific expertise or at least high-level knowledge of the specific changes being made.

If we wanted to move forward with c-neurons, we could come up with a careful plan to progressively decentralize them. Perhaps the security-sensitive c-neurons would be required to have a certain amount of DFINITY-appointed experts in them until a specified time in the future.

Starting with the roadmap seems like a better idea right now though. Not as much is at stake.

[justmythoughts]

I was just reading into the quote you linked, can you provide a bit more context on the ideas being passed around by Eva and the Growth team?