AGP discussion: An "Executive Branch" for the Aragon Network

[light]

This AGP is in the earliest stages of ideation. I’m starting a thread here to discuss before formally proposing in the AGPs repo.

Background and motivation

There have been several instances where AGPs have been proposed that specify an action to be done, but where the executor of the specified action either is not designated or is designated as the Aragon Association. For example, AGP-10 designated the Aragon Association to create a Community Funding DAO, and the draft “Prioritize Delegate Voting” AGP demands the Association to prioritize the development of delegate voting infrastructure despite the AA not having an internal development team of its own.

Indeed, AGP-1 even suggests that the AA is a kind of “executor of last resort” (my term) where it says:

If a vote on an AGP produces an Approved result, then the AGP will either be executed automatically by the Aragon Network or else dutifully executed by a manager designated in the AGP (or designated by the Association board if no manager is designated in the AGP).

AA Executive Director @stefanobernardi anticipated increasing demands on the AA to carry out the wishes of ANT holders in his comment related to AGP-10 execution:

AGP discussion: Community Funding DAO

I also think the responsibility for making sure the AGPs get executed should rest on the AGP submitter, while the AA retains the responsibility of getting it all approved, executed and making the necessary transfers in a timely manner.

The main reason for this is that the AA is not in control of the amount of things that are going to be asked of it in AGP proposals, and the team will be small, so I’d suggest, in order to make sure everything gets completed fast to minimize the load.

With this context, two things become clear:

1. It would be valuable to have someone available to execute AGPs that do not specify a manager / executor, so that ANT holders can approve funding for a project even if it is not yet clear who will fulfil the deliverables of the project. For example, the Prioritize Delegate Voting proposal. ANT holders should be able to say “We approve X amount of DAI for someone, anyone to deliver a working, secure Delegate Voting app” and for an executor to carry out those wishes.

2. The Aragon Association may not have the resources at this time to be the Executive Branch of the Aragon Network.

In response to these conclusions, I propose we:

1. Modify AGP-1 to more formally specify the Aragon Association as the “Executive Branch” of the Aragon Network, which will be responsible for dutifully executing AGPs that do not have a designated executor to the best of their abilities.

2. To require that for AGPs that do not have a clear executor or recipient of a Finance track proposal, that some portion of the approved funds going to the AA as an “Executive Fee” to compensate for the time it takes them to hire an appropriate team to build the app. This fee can be a flat rate, some % of the approved budget, or an amount negotiated by the AA and the proposal author - I’m open to whichever the AA and the community think is best for both sides. This may help fill the resource gap, and enable the AA to contract outside temp resources if necessary to execute a given AGP.

I’m interested in hearing feedback from the AA specifically if they are willing to take on this role more formally, and from anyone else who has questions or comments about this proposal. If the AA is unable/ unwilling to take on this role, but the community still finds such a role valuable, we can then work on figuring out an alternative (I have some ideas, but let’s discuss this option first.)

@stefanobernardi @LouisGrx

[stefanobernardi]

What an interesting discussion. Thanks so much for bringing this up @light.

It seems you’re not even considering to suggest that AGPs must have a designated executioner (who must also be willing to execute). That would be the easiest scenario, as voters can evaluate the abilities of the executioner while performing the vote.

It would also go a long way to reduce spammy AGPs, as the effort is shifted to the proposer. It’s too easy to say “oh we should do this, let’s create a vote” without having to think the execution through.

I tend to favor this line of thinking because with both 1 and 2 proposals you make, what worries me most is the coordination cost of those executions.
Getting compensated is not the issue - finding the time to hire someone is.

I think this could transform itself into a question of: what do we want the AA to be? Do we want it to be a nimble entity that exists today to make sure the treasury is used in accordance to AGPs and that “keeps things running”, or do we want it to be a much more structured and large entity? I fear the latter can go mimic current nation states - where Parliaments propose a million things and executive entities are inundated with work and rules at a level where execution quality plummets.

Thoughts?

[LouisGrx]

Hey @light, deep questions here! Adding my thoughts on top of Stefano’s

Ownership and accountability for AGPs

I’d like to challenge the premise of this proposal: that it is fine to receive AGPs with no specified executors.

While I haven’t done much research to understand wether it is possible, don’t you think AGPs without specified executors should be discouraged and at least remain exceptions?

Ownership and accountability are amazing tools to get things done. Specifying an executor a before proposal is approved by voters enforces healthy accountability on both the proposer and the executor.

I don’t really like a scenario where you have more and more proposals passing without a specified executor.

• It feels like the AGP process could be perverted, and increasingly become an instrument, feeding a soul-less execution arm.
• Most importantly in my eyes, it also feels like a way for token holders to come up with proposals and force them on the network. We probably want to think about this aspect carefully. I personally think that it would be too much power in the hands of ANT holders and that the current governance equilibrium holding between all stakeholders would be impacted significantly .

Personal conclusion: I’d rather forbid AGPs with non-specified executors or alternatively work on formally defining what is allowed.

About AA taking on the executor role

Other than discouraging AGPs without specified executors, here are the two main reasons I think AA shouldn’t develop internal resources to become an executor.

Minimize AA’s role in the Aragon Network

For now, AA intervenes anywhere things are not working/enforced autonomously: network coordination, due dilligences, treasury management, and real world operations related to those. In my opinion AA’s functions in the AN should be minimized and replaced by Network mechanisms as much as possible. So the idea in my eyes is to avoid adding new functions to AA including an executor arm.

Using and developing Aragon tools vs. using AA

Overall having solutions deployed at the level of the Aragon Network rather than AA is probably healthier for the long term. Therefore, if it is decided to allow non-specified AGPs, maybe we want think about some new processes to execute AGPs or use what we already have. I’m curious to hear about your other ideas (i.e. those not using AA as an executor)!

I can think of some already existing solutions and some we could create.

• Using already existing infrastructure like Flock and Nest to implement AGPs such as small projects or roadmap elements provided parties can agree before proposal is approved (Flock teams for roadmap elements, AA for Nest…).
• We have nascent research and processes to create bounties processes (CFDAO, TPS bounties DAO), I’d be happy to work on those to implement smaller AGPs and maybe even enable proposers to use a “bounties track” for execution?
• Should we research a specific bounties processes dedicated to AGPs execution?

With the above in mind maybe we could propose more formal tracks/options for the execution of “executor-less AGPs”

In the meantime, AA can ensure things don’t break

Until we have better tools and processes so things can ‘self-organize’ more, It is fine in my eyes that AA spends some resources on the execution and coordination of these -exceptional?- AGPs that have no specified executor. This probably only a medium term (6-12 months) solution?

Looking forward to hearing your ideas on the matter! Moreover everyone is highly welcome to think about the role of the AA in the Aragon Network

[Aaron]

If like to weigh in a little here as I drafted “Prioritize Delegate Voting”. Without the AGP process, As a community member and token holder, it’s not clear how I can influence the development priorities outside of forum posts and chats.

If the responsibility to execute always rests on the submitter, this closes down the only formal way for my voice to be heard.

However, after giving it more thought, I’m not sure this should be an AGP at all, there isn’t a clear track for it. On a related note, maybe there should be a method for formally signalling preferences.

Personal conclusion: I’d rather forbid AGPs with non-specified executors or alternatively work on formally defining what is allowed.

That being said I don’t think there should be a blanket ban on non-specific executioner AGPs, especially when there is no reasonable alternative. I do think formally defining what is allowed has a lot of merit. In the process beefing up signalling opportunities.

maybe even enable proposers to use a “bounties track” for execution?

This would be ideal IMO

[LouisGrx]

Hey there
I think roadmap prioritization by community/token holders happens each time a Flock team is voted in. This being said it makes a lot of sense to have a signalling tool for the community to have a continuous voice on the matter! Could be super useful as well for users to signal their needs and wants to flock teams.

Also think a ‘bounties track’ is worth studying!

[light]

@stefanobernardi @LouisGrx thanks for your comments / feedback on this!

After thinking about it a bit more, at least in the case of Finance proposals without an owner, it might make more sense to put them through the Nest process. As I understand it, the Nest process starts with someone (not necessarily the person/team chosen to complete a project) proposing a project to fund. Then if the AA approves the general concept, individuals or teams can begin submitting proposals to get funding and complete the project. Perhaps this could be a good way to direct these kinds of proposals to a different, established process where the author might still be able to get what they want?

To some of the comments/ questions:

I think this could transform itself into a question of: what do we want the AA to be? Do we want it to be a nimble entity that exists today to make sure the treasury is used in accordance to AGPs and that “keeps things running”, or do we want it to be a much more structured and large entity? I fear the latter can go mimic current nation states - where Parliaments propose a million things and executive entities are inundated with work and rules at a level where execution quality plummets.

I think it’s possible to head off this problem by capping the number of proposals shifted to the AA for execution, but can see where your concern is coming from. I don’t have many thoughts myself about what the AA could or should be, except to say that today it serves as a useful stand-in for the Aragon Network until the on-chain tools are more mature, and as a useful bridge to the “old” world that may be necessary even after the Network is live and fully operational. If the AA chooses not to accept the role of “executor of last resort” I could still see that falling to someone else if Nest or some other bounties process is insufficient.

Most importantly in my eyes, it also feels like a way for token holders to come up with proposals and force them on the network .

I think of ANT holders as the Aragon Network so to me this is contradictory. Edit: Meant the crossed-out comment only in a narrow sense, deleting so as to not get hung up on this.

In any case, AGPs cannot force anyone to do anything they have not already consented to. Am I misunderstanding your point here?

• Should we research a specific bounties processes dedicated to AGPs execution?

Per my comment about Nest at the beginning of this post, maybe the Nest process is good enough? And for the CFDAO we could implement an “RFP” convention that is similar, just for smaller proposals that are more appropriate for CFDAO funding than a larger Nest grant.

[LouisGrx]

After thinking about it a bit more, at least in the case of Finance proposals without an owner, it might make more sense to put them through the Nest process.

Yep Nest could work for some of them!

In any case, AGPs cannot force anyone to do anything they have not already consented to. Am I misunderstanding your point here?

If AA was in charge of executing executor-less AGPs, who would ‘be in charge’ of consenting to these AGPs when they go to the ballot?

Per my comment about Nest at the beginning of this post, maybe the Nest process is good enough? And for the CFDAO we could implement an “RFP” convention that is similar, just for smaller proposals that are more appropriate for CFDAO funding than a larger Nest grant.

Agree that for some proposals we could redirect them to these already existing solutions. On the other hand, the AGP process is a special process, probably more legitimate than any other in the Aragon Network, and ANT holders may still want to have the ability to create bounties with it…? Not sure there is a clear need for that though. (edit: I’d still personally discourage this track or any executor-less AGP, seems less efficient)