Creating some structure for our governance

Hi all! I’ve been thinking that perhaps a good way to go about our governance would be to create a clear schedule, where we debate specific topics for pre-defined periods of time.

So for example —placing arbitrary numbers just to illustrate — $UBI issuance could be adjusted once a year and we could have a 3 month period to work on it together. The process could start with discussions on the forum and sentiment gathering tools like pol.is, then have a few panels with vocal members of the community, economists and UBI proponents to continue deliberating, and only towards the end move to the proposal writing and voting phase.

After the voting is over, we can move to another topic, for example, registration requirements, where we all dive into sybil resistance together.

I think it makes sense for $UBI grants coming from the 4m pool to be ongoing, year round. But we could also fit them somehow in this schedule and have some process by which we can collectively elaborate on our priorities, even if the result we get at the end is not binding, and just for our own collective sense-making.

This would be somehow similar to how gitcoin grants work, where you have 4 rounds a year. Having some pre-defined structure can really help the community situate itself, and give more legitimacy to the governance overall. Just thinking out loud here, but would love to know what y’all think :slight_smile:

4 Likes

Seems like something reddit-ish would be good for this.

1 Like

Sounds like a cool idea. Would there be any problems with the rigidity of such a schedule if something of an urgent nature needs to be debated? Or would there be a process in place to insert such urgent matters into the timeline without breaking anything?

I’m also a fan of the sociocratic concept of ‘good enough for now’, where decision making isn’t too bogged down in search of perfection, but is iterated in a kind of agile manner to keep things moving. Not sure how that would fit in here though.

3 Likes

This is a great point @NTP! I definitely agree that the process should remain permissionless, so all members are able to submit proposals to the DAO at any point in time, to allow for that kind of spontaneous emergence and adaptability to changing circumstances. Otherwise its not a DAO! :slight_smile:

This would be more like an additional layer on top of it to allow for some degree of coordinated deliberation and deeper learning, by focusing on one specific topic for a given period of time. And goes without saying that such a process could only be instituted if it is voted on by the community.

1 Like

Cool @paulaberman . I admit I am still struggling to understand exactly how PoH prevents bad actors from subverting or corrupting voting. There seems to be no sanction which can be applied to stop someone (or possibly a group) doing this? Or am I missing something?

Hmm…@paulaberman I just read your Proof of Personhood piece on Coindesk, and I think my question was answered there. It’s not about identity so much as humanity I guess? Are you a real human, irrespective of your name and address etc?

1 Like

Well I think individuals can make their own schedule on what they want to work on. But a DAO-wise schedule would lower the DAO speed and make the governance even more bureaucratic (you’d have tons of debates about which topic would need to be considered).

1 Like

I think a simple quadratic vote once a year could solve the prioritization of top themes, no need for tons of debates there, as QV will give a pretty accurate signal and helps find common ground. This scheme is also not diminishing the agency of individuals in any way because they can still make their own schedule to work on what they want, and make any proposal to the DAO at any time.

I think it’s the very opposite: individual agency would be further enhanced because members would additionally be able to choose to deliberate and participate actively in a more coordinated effort if they want to do so, rather than just being restricted to raw voting.

Otoh leaving this process to self-organization would mean that only those who have the privilege of having a lot of time on their hands would be able to navigate through the cacophony of uncoordinated discussions. Meanwhile, the opportunities for real, collective learning and deliberation - which are key for a healthy democracy - would be significantly diminished.

1 Like

Exactly. Here’s a paper I co-authored on this subject, if you are interested in exploring this more in-depth: https://www.frontiersin.org/articles/10.3389/fbloc.2020.590171/full

1 Like

People may want to organize some specific period where they are focusing on one topic, but I don’t think it should be done DAO-wise.
From my experience “prioritization” in open organisation leads itself to a reduction of its resources as participants generally come to participate on one specific topic and not the organisation as a whole.
The risk is that someone will come with a project/idea and will be told “well this isn’t the priority at this moment” … and will promptly leave.

So if someone wants to organize some specific sessions on a specific focus, that’s cool. We may even have a vote on what some DAO workers should focus on.
But in my opinion, it should never be made in a such a way that it makes other discussions “off-topic” from some period of time and that it affect what people can focus on during their time not paid by the DAO.

My point is that we should only work on “positive” (“let’s talk about that”) prioritization but never on “negative” (“let’s move that for later as it’s off priority for now”) prioritization.

Maybe you could start a proposal giving more details on how you would see that working.

3 Likes

Yes, I agree these are super important concerns to keep in mind :slight_smile:

I can think about it in more detail and write a proposal at some point, but rather see if this can mature for a bit as an open discussion. Which goes back to my point on the importance of deliberation. Starting a discussion with a detailed proposal feels suboptimal to me.

That makes a huge amount of sense. Is there some way of using Agile voting methodology to keep things moving?

1 Like

Bit of a ramble but some of my thoughts on discussion/decision making:

I had been wondering about an idea of a ‘fast forum’ and a ‘slow forum’. I don’t know how formal this needs to be though, there will be a jumble of decisions that need to be made asap alongside longer conversations about the direction of the work. But I do like the idea we might all be playing with a concept at the same time which can give some focus and feedback loops for each other.

In the co-op I work with we are only a small team and have a similar process. As we are unable to break into smaller working groups we are forced pick something we are all interested in or think needs doing for a few months, rather than trying to do everything at once.

I also wonder if PoH will ultimately become multiple daos that work/compete with each other, if it is voted on by all of humanity then will there be some unresolvable divisions? It seems like a vote to fork away funds to another dao would be easier here than with other daos (e.g. If someone was to propose a vote to give the daos funds to the currently registered users, how many of the 1352 current voters do you think could be persuaded even though it is a bad idea? as more people join does this become harder or easier? populism can strike fast!).

Having “the way we do things around here” be slower, with longer periods of discussion and research may seem like unnecessary bureaucracy, but that process could offer great resilience to social threats.

another side is: perhaps we don’t want to accidentally create institutions/state government, perhaps we embrace the chaotic energy, move fast, test in prod. do we just have to try things and incrementally tweak the process as we find problems? does “deal with it as it comes” leave us open to unknown attack vectors? would a defined process just be an illusion of protection anyway?

Do we need a minimum HIP discussion period? or to say how long each vote should last for? Is there a distinction between the voting process and the discussion process or are they part of the same thing? do we only vote on disagreements or on any decision? Is the quorum 50% or higher? Do we include an abstain/opt-out option in votes to signal disapproval?
Do different types of decisions get voted in different ways? (e.g distribution of budget might be QV/ubi burning, and a binary choice 1p1v). Are the votes public or private (prevents collusion?).

I agree that we should have a “lets talk about this” approach for people to bring in ideas, but i’m unsure of how we go from that point to “lets vote about this” or “lets now do this”, and maybe those are longer, more deliberate conversations.

3 Likes

I’m here for “Decentralized Virtual Micronation” themed DAOs. :pray:

Our 1st problem is GAS for on/off chain coordinations but I’ll skip that part for scaling L2/PoH chain to straight talk about basic structures.

We don’t have to create big centralized structures around current form of 1 human = 1 vote governance. It should be more open and decentralized with self governing child DAOs.

  • 2+ verified humans can form/join child DAOs “orgs” with their own goals, rules, internal governance structures & voluntarily pay “taxes/donations” to parent DAO.

  • Child DAOs can join multiple alliances with other DAOs to push their agendas for PoH voting.

It’s loosely based 3 layers of decentralized federalism with bug fix.

  • Fed : 1 Human = 1 vote governance
  • State : Alliances of child DAOs
  • Local : Child DAOs

I don’t know how far this can go, I can pull some prototype codes if that’s going to explain this topic better. Basic idea is to experiment more and don’t end up with one/centralized governance protolol forever and ever for everyone. Only immutable core is 1 human = 1 vote.

4 Likes

I bellieve there’s a lot we can learn from Sociocracy here. The ‘good enough for now’ form of decision making, the determination not to get bogged down in consensus in order to be as efficient as possible. The DAO kind of takes care of the process foundations in a technical sense, but then surely there needs to be a layer of human interaction on top as an interface?