First of all, sorry for participating late in the conversation. It’s been quite overwhelming for me to keep up with all the most recent and important conversations happening in the Forum (I believe this is a shared feeling among other contributors within AA/AL too).
To put myself in context: I’m currently acting as co-CEO of Aragon Labs, the Swiss AG for-profit company that was set up between Aragon Association and the Vocdoni team after joining forces back in December 2020 with the objective of bringing decentralized voting tech to mainstream.
I believe that I speak on behalf of many AL team members when I say this:
We are happy to change the current setup if that can increase incentive alignment towards ANT, and we feel comfortable with switching to a non-profit legal entity that unifies all core contributors. Vocdoni’s roots have always been tied to a non-profit / public good ethos, and we’ve historically used “for-profit” entities as a vehicle for fund-raising only.
We welcome the initiative of decentralizing Aragon, and we believe it’s a step in the right direction. Considering the significance of the process, we feel that process should be safe to try from a governance perspective. A gradual roll-out with clear milestones would be our preferred approach so that we can fail small (I’m sure there will be failures, and we should welcome them. We just need them not to be fatal). On a personal note, I fully trust in the capacity and outstanding work that community members such as @lee0007, @fartunov, @AClay, @daniel-ospina@b3n are spearheading, and I hope I can contribute positively to this conversation too.
We believe that the best way we can help bring value to Aragon is to ship innovative Product and Tech aligned with Aragon’s Manifesto, which we are using as our MVV. We are now focusing on the short term to reach operational excellence so we can ship at the speed that the industry requires, and we look forward to increase our engagement within the AN DAO with better communication and coordination with the community.
We welcome collaboration with other teams who also want to build products and tech that supports Aragon’s mission (1Hive, EVMCrispr, others).
Overall, feeling positive and excited towards this new phase for Aragon.
Can you share more info please on this? in which jurisdiction(s)? To confirm, if ANT holders controlled the General Assembly of the AA and it’s funds via an appointed committee, assuming there was not a large scale transfer to the DAO, do you see that as better from a security risk standpoint?
There are several proposals for enhancing the utility of ANT that are related to securing components of the decentralized infrastructure of the Aragon Stack. If an increased utility is provided, wouldn’t this strengthen the utility classification of ANT?
Why would granting the DAO power over the treasury modify the nature of the token towards a security? Shouldn’t this be the opposite, as we’re advancing towards “sufficient decentralization”?
The law which is more important for us is the Swiss one. And of course @ronald_k has the last word on this topic.
The essence of the problem is that with the proposed setting the tokenholders could theoretically distribute the treasury among themselves, as a company that distributes the dividends or that goes into a liquidation procedure. This would make ANT an investment token in my opinion. Given the peculiar features of Aragon (big treasury compared to the marketcap), this is something that we have to bear in mind with particular attention.
This problem could be avoided with particular limitations concerning the DAO governance. But it is something that requires a proper assessment.
On this, the current charter already has a requirement for finance proposals to include a justification of why the proposal would be likely to increase the number of active Aragon DAOs. As such, it’s unlikely that anyone would be able to make a finance proposal for an ANT holder dividend payment and get it passed.
If that’s insufficient, we could simply add an immutable clause to the new charter that prohibits dividend payments to ANT holders.
This is a joint statement by all of the Aragon Research team.
We fully support the idea of being part of a DAO, funds included. This is why we joined Aragon.
However, we think this proposal must be amended to enable an in-depth debate on such a sensitive topic and thus requires:
The information provided must be more detailed to show the suitability of the proposal to achieve Aragon’s mission.
The time schedule is too tight to make the required amends and allocate enough time to include all stakeholders.
Specific comments on financial aspects:
Given the total amount of funds, the transfer of assets should be implemented gradually through a number of iterations, according to a clear roadmap and subject to clear criteria. These criteria should include a measure of the maturity of the governance of the AN DAO.
Specific comments on governance aspects:
We think that the proposed quorum and thresholds are too low.
As is the case with most DAOs, the AN DAO suffers from a severe lack of engagement by ANT token holders (as already pointed out by Brent in a previous comment). As the AN DAO will eventually be managing a considerable treasury, it is crucial that it achieves functional maturity in terms of both governance structure and rules, as well as token holder engagement. To achieve this, the charter will need to be amended, most likely in several iterations. We may also consider adding some type of incentive for participating in governance and voting. This will take time. We note from empirical observations that delegated voting does not increase voter participation if the distribution is already highly non-uniform (see for example https://compound.finance/governance) or there is a cost to token delegation.
The iterative process of improving the AN DAOs governance can be used to gradually add innovative features to the AN DAO (for example weighting votes for accounts linked to an identity protocol, anonymous voting using zk proofs, quadratic voting).
In order to become natively delegatable, the ANT ERC20 contract would have to be updated by adding the [ERC20Votes extension](ERC 20 - OpenZeppelin Docs). This would open up the possibility to use it in bribery protocols/vote lending platforms.
Vote lending / bribery is forbidden per our charter (see chapter 4 “The Community Guidelines”, “Our standards”), but there is no technical measure to prevent the usage of bribery/vote lending protocols.
The other option is to separate the delegation logic from the token so that it is part of a governance module, which might require a security audit.
Questions to the proposer:
Is the proposer willing to agree to a gradual transfer of the funds given the risks pointed out, or are there arguments speaking against this?
What is the motivation behind/benefit of using delegated voting given the concerns mentioned before?
How is the accounting/operational side intended to work? How will the particularities of each team be taken into account? (e.g. in research, the working cycles are usually long, the outcomes are frequently intangible, and failures are not uncommon).
I am in support of any actions taken to decentralize our governance sooner than later, and I am glad that this proposal speeds up our collective efforts.
However, we must be careful that the speed and effort does not lead to undesirable, intermediate centralization along the way.
For this reason and with both of my job roles as governance researcher and engineer, I am concerned with weaknesses of delegate voting that you will also read in the joint statement of the research team. The associated post is unfortunately not yet visible her because the posting account was automatically placed on hold, probably as a spam protection measure.
Here are my personal considerations:
If the delegation is achieved by extending the ERC-20 token with the ERC20Votes extension (Compound-type delegated voting), the usage of bribery/vote lending protocols increases the risk for governance attacks (as seen recently, for example, in Compound) and is hard to prevent technically. Furthermore, it should be noted that bribery/vote lending is prohibited by our charter (see ‘Community Guidelines’ and ‘Our Standards’ therein).
Examples of unacceptable behaviour include:
Selling or renting votes.
Assuming that people use ANT on vote lending/bribery protocols, this can centralize the decision-making process. In this context, it should be mentioned that bribery can generally also happen outside of bribery protocols, but coordination is usually much harder in this case. If delegation is achieved by additional means outside the ERC-20 token, it becomes technically more difficult to use them in bribery/vote lending protocols. A mitigation might be to deny known bribery/vote lending protocol / vote staking pools addresses from voting. All the connected issues should be carefully considered.
I would like to hear the assessment of the Tech Committee Sub-DAO (@nivida, @voronchuk, @p4u) on the points above in line with their responsibilities outlined in the charter (chapter 6, section 3.a.i.1 to 4).
i. In particular, the Tech Committee DAO has the following responsibilities:
Review technical proposals in the Main DAO and Sub-DAOss to assess them from a technical risk perspective.
Remove technical proposals in the Main DAO and Sub-DAOs that represent a material technical risk to the project.
Approve technical proposals they believe would be beneficial to the Aragon project and DO NOT require a 3rd party technical security audit due to being low risk.
Suspend technical proposals they believe would be beneficial to the Aragon project, pending the completion of a 3rd party technical security audit.
However, I strongly agree that we shouldn’t find technical excuses to not decentralize our governance, and we could, for example, put conviction and TAO voting up for discussion, that have more security measures in place.
HOW ARE VOTING MARKETS CREATED?!
Paladin can integrate into any governance system that utilizes delegation mechanisms. As of now the team is hand picking select protocols with the goal of later simplifying this process.
VOTER EXTRACTABLE VALUE (VEV)
VEV is the protocol-agnostic use case for voting from Bribe. VEV allows anyone to lease the voting power of their governance tokens by staking them in the Bribe pool for their specified DAO. Highly incentivized users then bid on specific governance decisions they want to approve or reject related to the protocol using the power of the staked votes.
Yes, that’s by design in all token-weighted voting systems. But note that anyone in the community could acquire delegated voting power through merit and reputation.
Given that ANT is freely tradable, the distribution of token-weighted voting power can’t be arbitrarily forced. An argument can certainly be made that a higher distribution is better but I would also claim that, in practice (assuming similar objectives, first and foremost the broad adoption of Aragon tools), the distribution of decision-making power is of secondary importance compared to the quality of decision-making. That’s why it’s crucial to have an open proposal process in which the best ideas can surface and be voted upon, ideally by a sufficiently diverse community of delegates who vote with the best interests of the DAO as a whole in mind.
I’m not sure I understand why contributors would lose motivation and knowledge if the DAO treasury is governed through a delegated voting system. It’s standard practice in many DAOs out there to use tokenized voting to ratify budgetary decisions. Indeed, it is one of the most essential features of DAOs.
I don’t think an intermediate step is necessary but setting up both the technical and social infrastructure may certainly take some time.
No, the whole treasury would be managed through a delegated voting system but, as a safety measure, there would be an annual spending cap of 10% of the starting value of the treasury. That number can be changed to whatever feels more appropriate, or according to the changing needs of the DAO.
I don’t think it’s rational to expect that the vast majority of token holders will participate actively in the day-to-day governance process. The current experience with DAOs shows a strong tendency towards professionalization and delegation. Also, malicious and/or highly contentious votes are by far the exception rather than the rule, which supports the hypothesis that, in the vast majority of cases, higher turnout wouldn’t actually materially affect the outcome. That said, both direct and delegated voting should be made as easy as possible so that, whenever the need arises, the DAO can mobilize the decision-making power of its members as a whole.
I definitely think Aragon DAOs should have a delegation feature but I’m not the right person to comment on how it’s best implemented given the current product design and roadmap.
How would you describe the purpose of the Association vis-à-vis the Aragon DAO as a whole and its mission to be a leading provider of DAO tooling?
I think it’s important to consider all relevant legal implications and inform the DAO’s members of the various risks and best ways to mitigate these risks. That said, the vast majority of DAOs are operating without a legal wrapper and instead have various functions/contributors housed in separate legal entities that handle compliance and protection for their members. This includes foundations for handling IP and private companies that employ DAO contributors or serve as delegates. While it is certainly experimental/innovative (the whole point of DAOs is organizational innovation), I don’t think it’s fair to say that these structures are completely untested.
Any organization that involves more than a single person can be described as decentralized to a degree. In the context of DAOs as a novel form of organization, however, decentralization usually assumes independence from a single organization based in a particular legal jurisdiction. But of course, that doesn’t exclude a whole range of such entities contributing to the DAO in various ways.
That’s a very different path from the one envisioned by the current proposal. If there’s broad enough support for it, though, ANT holders could certainly be given a choice between the two options (or even additional ones, should they emerge).
EDIT: Based on what @joeycharlesworth wrote below, one option doesn’t necessarily exclude the other, although it would certainly be a very counterintuitive thing to do in light of how most DAOs have aspired to structure themselves in the past.
One does not exclude the other but the current proposal is focused specifically on distributing the control over the treasury through a delegated voting system which is viewed as one of the core functions of DAO governance.
Do you or @ronald_k have prominent examples of this type of DAO practice? It is my impression that DAOs generally prefer the exact opposite path, namely of avoiding organizing themselves inside a single legal entity, which is better aligned with the value proposition of having a DAO in the first place.
I think we’d be the first Swiss Association to do this but will defer to @ronald_k if he has other known examples. To be clear, what I’m proposing is decentralisation of the treasury to AN DAO (controlled by ANT holders) AND having the Aragon Association controlled by ANT holders. The result is obviously more decentralised than having the Aragon Association continue to control access to key resources other than treasury, without ANT holders having a voice within it.
Got it. I think a good case study here might be the progressive decentralization of MakerDAO which included the Maker Foundation handing over its treasury to the DAO and the establishment of a DAO-funded Dai Foundation to safeguard the community’s IP. I’m not saying that this is exactly the right path for Aragon but it’s a good precedent to learn from.
The choice between a minimum viable design vs. trying to anticipate all future challenges depends on the objective. The objective of the current proposal is to distribute control over the DAO treasury through a delegated voting system. However, there may be other objectives that need to be achieved prior to taking this step (e.g. revising the Charter and expanding the capabilities of Aragon tooling).
I would definitely second that.
One does not exclude the other and I don’t see distributing control over the treasury in any way hampering the effort to build up the off-chain social capital and governance capabilities of the DAO.
The proposed timeline needs to be reviewed carefully after we’ve established the list of steps that MUST be completed prior to implementing Steps 2 and 3 of the original proposal. I agree that this includes establishing MVV (perhaps as part of a general revision of the Charter) and figuring out the basics of legal risk mitigation in the context of delegation.
With regards to technical infrastructure, it sounds like the necessary tooling already exists, but there may be a need to port over to a different solution later down the road once delegation is implemented in the current version of the Aragon software. On this topic, I trust the judgement of those most familiar with the technology.
I like the idea of a working group to finalize the proposal and the timeline.
Correct, the proposal assumes token-weighted voting as the default governance mechanism. However, delegation is viewed as a way to empower any member of the DAO based on merit and reputation. In the future, the DAO may implement other governance mechanisms not tied to token holdings.
As @Joan_Arus pointed out, it would add unnecessary complexity/distraction to the voting process by triggering ANT market action. However, it is certainly something that the DAO can decide to implement in the future.