Hey @Lorimer are you trying to argue that @Cris.MntYetti didn’t post exactly what he meant to post and that it didn’t accurately capture his thoughts on the subject? Do you realize that LLM are considered a new tool to help streamline and improve communication? I mean, you can still walk over to a neighbors house to have a one on one conversation, but after the telephone was invented it also became acceptable practice to pick it up and talk from different locations. To each their own on how they choose to communicate. What is important is that Cris gave you his exact thoughts on this subject. He expressed genuine concerns and your rebuttal is to criticize his communication style. You turned much of the Q-QUORUM thread into ad hominem attacks and now you seem to be doing the same to this thread. Can we please keep this thread clean and just present your own case for this change so that others can read and respond objectively?
Addressing those who Alex tagged: I’m just looking for feedback from people with the necessary knowledge to confirm or change my thesis on this proposal. The post has been merely structured using the LLM based on my own interpretation and input. I would appreciate any kind of feedback on my doubts.
Unrelated
Now Alex, I try to remain objective and on topic and you have done nothing but continued to roll back with personal attacks. I don’t need your drawings to understand things, I’m not in kindergarden.
Yes, I do use LLMs, both for gathering information and polishing my writing. I use LLMs to format the text in a more polished version than my dyslexic input. For your information I’ve spent around 12+ hours to document myself on this proposal and I asked politely to be corrected if something was misunderstood. You see, even though I use LLMs, I also look for confirmation from knowledgeable people on things I might lack before drawing my conclusions. You have done nothing in this regard, yet you have dedicated quite the effort to make me look bad or like some kind of impostor. Congratulations, that really says a lot.
I work with CodeGov, yes, but first of all I’ve been an ICP investor since 2021 and have done countless hours of research. I don’t need your data, I’m very aware of the situation. Clearly it is you who doesn’t understand what I mean when I say I want VP to be decentralised. Would you like me to make a drawing for you? Your proposal impacts MY investment, i’m doing my due diligence on MY investment and your only argumentation for now has been that I work with CodeGov, hence I’m biased.
Your D-Quorum neuron makes governance redundant and that’s a fact. You want people to vote for a council that votes for proposals that people can already vote on or delegate to others. Your problem is with the fact that the majority of neurons controlling NNS and SNS are managed in a centralized manner, and I see no problem in that. Neurons can be individuals or they can be collectives that manage them as they please. Contributing to a decentralized environment does not mean that everyone has to do it in a decentralized way, otherwise Dfinity itself should be under the control of the NNS. You want a DAO controlled neuron to counteract? It’s great, do it, I praised your effort, but putting it under the control of the same DAO it contributes to is pointless. SNS projects are not TAGGR that has no following mechanisms where it would make sense to have a DAO governed council that acts in the name of the DAO. SNS projects already have those councils, and it’s thanks to people having the ability to follow whoever they feel represented by the most. I guarantee you that those asking for the dev neuron do so for a specific reason, being stupid as you called them is not it. Congrats on that line as well btw.
On a side note, if CodeGov secured those grants is because it has a multi-year track record of being a reliable entity in the ICP ecosystem. You were part of CodeGov as well, you took those grants too for the record, besides getting one for yourself. I had a lot of respect for you Alex, your technical skills are not up for debate. You lost it the moment I came up with a couple of arguments that do not align with yours. I tried to remain on topic and to distance myself from your personal problems with CodeGov, you have done nothing but dragging me in. I’m done talking to you, you clearly can not help me understand the benefits of your proposal for me to make an informed decision.
This has been really unrelated, and frankly quite unnecessary.
@Lorimer your power law concerns will be addressed when DFINITY implements the following items, which are on the roadmap and being worked on actively.
- implement private/public neurons
- implement periodic confirmation of neuron followees
- implement known neuron registration enhancements including registering for specific topics and deregistering known neurons that are not active
- displaying a list of registered known neurons for each topic when users are trying to decide who to follow for a specific topic in the NNS dApp
- displaying warnings in the NNS dApp regarding topics that are not included in the All Topics catch all category
- displaying warning in the NNS dApp when people have neurons that are eligible to vote, but are not actually receiving rewards for voting on specific proposal topics for whatever reason (Followees not configured, Followees not doing their job, neuron owner not voting manually, etc)
- There may be value in considering removing topics from the All Topics catch all when there are known neurons that have registered for that topic. I haven’t seen this proposed anywhere, but it would make sense as a mechanism to encourage people to select known neurons for topics that have the potential to be decentralized because people and organizations have stepped up to present themselves as a valid option. I’m sure DFINITY will register for each topic individually, so they can still always choose to follow DFINITY.
- offering NNS sourced incentives for people and organizations to perform the work of proposal reviews
These items above still don’t address the concern that has been expressed previously about why neuron owners would choose to follow someone other than the dev team, which in the case of the NNS is DFINITY. Regardless, I feel very confident in saying that nobody in the ICP ecosystem believes more strongly in decentralization principles than folks that work at DFINITY. I believe that their goal is to carefully and intentionally move the ICP community toward decentralized governance on the NNS as well as SNS projects. Decentralization is a spectrum that will follow a transition measured in years and we are already on a trajectory that advances decentralization. DFINITY has enabled this to occur with the changes that they have made since genesis and all signs point to that continuing.
I was a strong advocate for offering grants to many people and organizations with the Grants for Voting Neurons initiative. The program is currently underfunded and I believe everyone who put themselves forward for the role should have been given the opportunity. However, that is not what was scoped and it is not going to prevent CodeGov from continuing to do our part to help advance decentralization as well. We have a team that I believe is well suited for all the grants we received and we have spent a year and a half demonstrating the capacity for these technical reviews. Our application was put before the NNS on every grant that was offered and we competed with all the other applicants on equal footing and based on the current state of decentralized NNS governance. It’s not perfect, but it did move NNS governance decentralization forward. More changes are planned including those listed above that will continue to advance decentralization.
So if your goal behind changing the core principals of NNS and SNS governance with this Manage_SNS_Controlled_Neuron method is to give other neurons a better chance to become known and recognized as credible Followees, then I believe that will already be achieved with these ideas that have already been discussed extensively in other threads and are already on the DFINITY roadmap. I do not believe any individual neuron should be given preferential treatment by being controlled by the entire DAO that it is meant to serve. All neurons that are contributing to the governance of the DAO should be controlled by individuals or smaller groups of individuals within the DAO.
I meant the words that I wrote (nothing more and nothing less). There have been no ad hominem attacks (none that I’ve meant as such), I don’t know what’s being referring to. Similarly, I’ve not called anyone stupid @Cris.MntYetti 
I apologise for anything that has been received that way - genuinely. I also apologise if I have hurt anyone’s feelings with any of my statements or observations (including the LLM itself 
). I won’t get into the debate about overuse/overreliance on LLMs, or the problems they can introduce for effective public debate. A topic for another thread.
I think my attempts to highlight the problems that a DAO-controlled neuron solve are being interpreted as attacks (and I can see that perspective). The trouble is that the problem that a DAO-controlled neuron solves is one of dangerously overcentralised DAOs (mostly due to complacent followers, and not enough reviewers stepping up to the plate). In order for me to highlight the need for this solution, I need to describe the problems.
CodeGov currently has humongous VP in WaterNeuron, particularly for the NNS proposal topic (the most important one). In other words, in principal a single person can instantly cause the WaterNeuron neuron to vote on any NNS topic (a person who is also doxed, which is even more problematic given the potential for hijacking). Given that 2 members of CodeGov also happens to be the only opponents of this DAO-controlled neuron idea (so far), my attempts to describe the need for this probably look like attempts to belittle the people I’m debating with. That’s regrettable, and I’m sorry that I have come across that way - really.
I haven’t got time to fully read and respond to the other comments this moment (I’m sure I’ve already done so elsewhere, but there’s a lot to read). Nevertheless I’ll try to make time for this later this evening or tomorrow. Thanks.
This is because the WTN dev team follows CodeGov WTN on the NNS topic.
We created the vote relay app to replicate votes from the CodeGov NNS neuron to the CodeGovWTN neuron. That app is available for anyone to use or to fork and create their own canister to use. The WTN dev team could easily choose to set up a vote relay app for any other NNS/WTN pair. I could do the same and so can you. Then the WTN dev team could choose to follow up to 15 neurons on the NNS topic. If they really wanted to get fancy about voting for a variety of people on then NNS topics, then they could create a new NNS neuron and set a variety of Followees for each individual topic and then create a vote relay that they follow.
I really don’t care who they choose to follow. I didn’t ask them to follow CodeGov and I have asked them to stop following CodeGov on at least 2 occasions. It’s their choice who they follow, but I wouldn’t be surprised if they follow CodeGov because they know we are reliable at always voting and we do it by seeking credible and intentional Followees for all topics and every proposal.
Even if I have the ability to cast manual votes for CodeGov neurons today, I do it only in situations where the configured Followees cannot reach consensus before the end of the voting period for some reason. I don’t need anyone to trust me. I’m not asking for anyone’s vote. This is simply my policy at CodeGov and if someone doesn’t like it then they are welcome to follow someone else. There are plenty of other choices today and there will be even more credible choices in the future. To each their own. Follow whoever you want. Just don’t try to make an entire DAO vote on Followee configurations for a specific known neuron when all we should be required to do is make our own choices on who to follow with our own neurons.
I think it will still take multiple months until we get there.
There are quite a few things on the roadmap that were requested and are eagerly awaited for by multiple SNS DAOs and therefore are prioritised first. This includes easier SNS framework upgrades (WIP), easier following by introducing SNS topics (under design) and a concept of named neurons, large wasm support, easier handling of treasury.
As there is some way to achieve what you would like to do (using an extra canister) and as there seem to be different opinions on whether this should be done at all, how about we take up this discussion again in the end of winter?
We could then take more time to discuss the pros / cons, also including new information that we learned from new features such as the named neurons.
@lara this is exciting to know that these things will be implemented for SNS projects by the end of winter. That sounds like amazing progress.
Don’t want to promise this timeline just yet 
I think by the end of the year we will have made great progress on simplifying SNS upgrades and at least have a good plan for SNS topics and critical proposals. Let’s see where we are by the end of winter with all of it, but even if not all of it is finished we will have more information!
Also excited to see all of what is coming and how this will be used by the SNSs!
Hey @diegop @Ang these two responses were flagged as off topic by someone in the community, but they are very relevant given the prior message. Will you please unhide them?
https://forum.dfinity.org/t/sns-generic-nervous-system-function-restrictions/36239/21?u=wpb
https://forum.dfinity.org/t/sns-generic-nervous-system-function-restrictions/36239/22?u=wpb
I agree, they’re relevant messages. My message was also hidden, and I had to edit it (slightly) in order for it to become unhidden. Not sure why it was flagged.
Thanks Lara, that all sounds good and I’m looking forward to what’s coming, particularly known neuron functionality for SNSs.
Almost, but not quite 
The crux of the idea is that when a new user joins a particular SNS (an SNS that has decided to adopt a DAO-controlled neuron), that user then has a 3rd option about how to approach voting:
- Vote manually
- Ask who the best neuron to follow is (usually the one that already has the most followers, which is therefore self-reinforcing, leading to undesirable centralisation)
- Donate your voting power (by following a DAO-controlled neuron) so that the actively voting members of the DAO determine how that VP is distributed to aid decentralisation
By instead moving this responsibility to a separate canister, governed by a separate/smaller group of people, you’re significantly reducing the number of people that can have a say in how the DAO should distribute VP that has been given up by users that don’t know (or have no preference) about who to follow. That may be a fine approach for some SNSs. I’m not suggesting a silver bullet (I think that’s where some wires have possibly been crossed). Ultimately, there’s no reason that any of these approaches need to be mutually exclusive of each other.
As an interesting case study, take the recent NNS grants for voting neurons elections. Would it have been reasonable to not allow the entire NNS to vote on which NNS reviewers should be regarded as preferential on specific topics? Would it have been preferential to instead hold a vote via a different governance mechanism, through a different canister, governed by a different much smaller set of people?
At the end of the day this thread is about requesting the functionality for an SNS to decide what to do for themselves in this respect. If they own a neuron, they should be able to manage it (including dissolving it). I’m looking forward to this funtionality arriving.
The WTN D-QUORUM neuron is c0b121b52cca921791b87350244029a9d128883e22576b528f2214489a1d3384
It is controlled by no other prinicipal than the WTN governance canister. There is no way for the DAO (or anyone) to manage and/or dissolve and disperse this neuron without a manage_SNS_controlled_neuron proposal.
This sounds like a nice timeline to me, thanks @Lara. I’d also like to thank the numerous people who have provided useful criticism, asked questions, shared ideas and shown their support and positive sentiment for this idea on numerous channels and DMs - including @sat, @rem.codes, @EnzoPlayer0ne, @1eo, @wpb, @Cris.MntYetti and various others who’s handles I’m not aware of on this forum.
You’ve all helped to flesh out some further details regarding setting a standardised set of principles that a D-QUORUM neuron should conform to. I’m aware of numerous individuals who are interested in setting up a D-QUORUM neuron on other SNSs, so a standardised set of principles to follow will be useful. I’ll plan to collate these and post them on a dedicated thread at some point soon.
Thank you for your feedback Lara. Looking forward to seeing the effects of these new features from the roadmap once they’re implemented. Especially the SNS topics and known neurons concept, it’s quite cumbersome to manage your following settings at the moment.
Hi @lara, now that we’re a couple of months down the line are you able to provide an update on this? @Mico has recently put forward other interesting use cases for SNS-controlled neurons.
This is about facilitating SNSs to set up their own governance preferences, rather than forcing a particular stance on any SNS
Cross posting here for visibility since @Lorimer continues to push this idea.
I am still opposed to the idea for the reasons stated below and in the original post.
Below is the response (also linked here) provided by @lara to this concern raised above…
While these comments were made in a different forum post, they are based on conversations that were held in this forum topic. The history of this conversation can now be found in this thread.
Compared to this request, we still have a few ongoing things like introducing SNS topics, enable governing dapps with large wasms (we are working on those), known neurons, more automated upgrades, etc.
As these have been requested by quite some SNSs, in my current view these things still deserve higher priority.
Can you remind me why doing this via an extra canister does not work? I start thinking that for a few capabilities having such a more modular approach and doing some things outside the main SNS framework can actually be beneficial: it can then be reused by others and integrated with the SNSs that need it while not complicating the SNS canisters themselves. It would also probably allow you to move much faster. (This is something that I wasn’t only thinking for this request but also another idea lately)
That’s understandable. When you can make some time, I think this would be a very minimal effort change to introduce. As a community member that’s active in reviewing SNS dapp canister code (I don’t think there are many of these), I can tell you from first hand experience that this is sorely needed.
Another angle to come at this from is that if a DAO is capable of owning a neuron, why shouldn’t the DAO be capable of controlling that neuron or deciding to dissolve it (particularly when it’s their native token)? Note that there are no restrictions on an SNS DAO controlling a neuron that’s native to another SNS.
Other WTN DAO members have also mentioned interesting use cases for a native neuron owned by the DAO (cc @Mico).
I will continue to argue that an entire SNS should never be required to vote on how one of its own neurons is configured with Followees.
In my opinion, a SNS should not be capable of owning one of its own neurons. The fact that someone can set a SNS governance canister as the controller of a neuron in its own SNS ecosystem is the problem and should be resolved by removing that capability.
I see no problem with this capability. An SNS should be capable of owning and controlling neurons of another SNS ecosystem or the NNS.
Hey @Mico would you please elaborate on the use cases where you believe an SNS should be capable of owning and controlling a neuron in its own SNS ecosystem? I don’t believe I’ve heard you describe these scenarios, yet I’ve seen Alex claim you have identified scenarios a couple of times. Why would an SNS owning a native neuron be preferrable over a canister controlled neuron where people in the SNS must opt in to participation of the functions designed for that canister? Why is it necessary to force the entire SNS to participate in the control of the neuron when the neuron is not their own personal investment?
Hey @lara is it documented anywhere why the SNS framework already excludes the SNS governance canister controlling a neuron within its own SNS? I assume this was intentional and I’d like to know why. I agree with this approach, but it may be important to clarify in case there really is going to be some future deliberation on the idea where restrictions may change and the feature request becomes part of the SNS framework. I’m curious why DFINITY is not already shooting down this idea given that restrictions are in place for a reason. If the idea has legs that DFINITY would seriously consider supporting, then I’d like to know sooner than later.
To clarify Wenzel, there are no security concerns. You’re asking to restrict the options made available to the DAO for no justifiable reason. The amount of work required to enable this feature is also extremely small.
Trying to avoid governance overreach is more than enough justification. Neurons are supposed to be independent, not state sponsored. Decentralization is achieved by having many independent neurons voting according to their own best interest so in aggregate they reach a decentralized decision. State sponsored neurons completely defeat the purpose of independent neurons. I don’t think it is appropriate for everyone in an SNS to involuntarily be required to vote for Followees for somebody else’s neuron. This is a principal that is worth standing up for until the bitter end. I hope we never have to bring this to an actual vote and it makes me nervous that DFINITY hasn’t shut it down already given that is completely changes the fundamentals of neuron voting and liquid democracy as we know it.
This is an interesting stance. Are you able to show your workings? Governance overreach is the current status quo. Just look at the follower-inclusive-VP of the most powerful neuron or two on any SNS.
Could you explain this further? Neurons are designed to be capable of having no explicit controller (‘community-controlled’) and there are also already many cross-SNS DAO-controlled neurons.
Why would that be the case? In any case, do you prefer this priviledge resting in the hands of an individual?
Yes. That’s not where we are, and if you track the stats, it’s not even where we’re going.
There’s no way a neuron like D-QUORUM would lead to worse centralisation than we have now. It distributes consentually delegated VP equally among established neurons.
Have you ever looked at a graph of the distribution of follower-based-VP? Would you like me to show you, in addition to showing you results after simulating the presence of a functional D-QUORUM neuron (under pessimistic, conservative and optimistic conditions)?