Revisiting the DAO Structure

Proposals
1

This is not a proposal as-is, but I think it can lead the way to a proposal. So I chose the governance category for it. If it’s not fit for this category, mods should move it where it is due. The information I use for my views are based off the info readily available in POKT docs.

SUMMARY

Current DAO Governance consists of users who pass the requirements and objectives stated here.

To give an idea, some of the things to do in order to get a DAO vote are:

  • Submit a community-orientated PEP and receive a POKT grant from the DAO.

  • Tweet about Pocket in a meaningful way, retweeted by 3 Verified Shepherd’s or Pocket core team members.

  • Create a Pocket meme and get at least 5 dab reacts from Verified Shepherd’s or Pocket core team members.

These are just examples. But what they show is that inorder to become a DAO member and vote on critical proposals concerning POKT, you have to be good with social media and well connected enough to get DAO members or team members to notice you or be a very good coder.

I believe the current DAO structure should be revamped. Here are some of my reasons:

-Making team members gatekeepers of potential new DAO candidates is a big red flag of centralization. And currently, the docs suggest they are.

-DAO members do not necessarily carry any risks associated with their voting powers. Since they do not have to carry risk holding POKT in order to vote, they can vote to negatively affect the price to buy back in at times. Or if they were to sell their POKTs before a bullrun, they can simply act against the interest of the protocol due to pure anger.

-DAO members carry the value of a single vote without any ongoing costs, meaning in theory that they can simply sell their votes for personal gain. They do not need to hold any POKT or have a stake in the protocol to retain their voting powers. They have infinite value without recurring risk costs typically associated with holding governance tokens to vote.

-After a potential DAO member completes every mission or trophy, they still need 1/3 of existing DAO members to vote to get in. This means any rival to POKT, can try to damage the protocol by snowballing DAO members once they manage to get a few in, since the largest barrier to getting in the DAO is getting the DAO members to do things for you.

So this means that the POKT DAO is simply being gatekeeped by a select few. I’m sure these are well intentioned people. But no one can vouch for the future. I have seen developers or teams constantly asking the DAO members for votes on their grants and who’s to say they can’t be bribed?

If the DAO proposals were voted based on the stake a user has with the protocol, including but not limited to token amount and node amount, a vote would be significantly harder to enforce via bribe. But currently we can see that some proposals pass with 3-4 votes only. Not that hard to bribe or contact these people. (In a hypothetical scenario)

I think the current DAO voting system has not yet shown it’s vulnerabilities, but it definitely will and it might be too late to change it when that day comes where interests diverge.

I have not yet seen any plans to change this structure. I simply go off the information available on POKT Docs as any investor would do before putting their money in POKT. From an investor viewpoint, I fail to see how the DAO members chosen as described in the docs can have the projects’ best interest at all times given the reasons I previously stated. Someone who has just the right skills and time but no real stakes in POKT can have a say in the future of the network while 7 figure investors can’t… It may seem cypherpunky, but it is simply not logical.

I propose discussing the pros and cons of the current system and replacing it with a new system that does not give infinite voting power and value, free of continous risk over completing a list of trophies.

2

This is literally the formula for pay to play. If your desire is to give majority control to the whales, this is how you do it.

Anyone in the current DAO structure can traverse the voter path with proof of knowledge and proof of service. It’s the most equitable system possible.

I strongly disagree with moving to a token based voting system.

2 Likes
3

I cannot speak to where this topic should be placed; a mod will determine this.

What you write could become the material for the "Motivation/Rationale"section of a proposal. What is being proposed as an alternative to current system is still blank. “Discussing the pros and cons of the current system” simply doesnt cut it as an actionable item to be voted on.

So perhaps resturcture this into: “Replace Trophy system with X”
Flesh out X
Give rationale why X is better than current.
Answer the inevitable objections to X

1 Like
4

I am none of these things yet have managed to earn a vote.

I do agree that perhaps there should be additional pathways to earn a vote, (Tokenomics is one that immediately comes to mind), but strongly disagree with the token based voting system you suggest.

3 Likes
5

The current DAO structure does not have checks and balances and it has no mechanism to reflect the consequences of bad decisions on DAO voters as financial loss if they don’t hold any POKT.

Crypto is a fast scene and anyone who invests the time and work to complete these trophies gain a vote forever and that power carries on with them even if they move to other things.

The current mechanism only requires initial work to be put in to get a DAO vote, and it’s not easy work I admit but after that you really do not carry any risk and still vote on the protocol even if you have nothing to do with POKT anymore.

In long term, this can effectively mean that a protocol with hundreds of million dollars at stake can be governed by people who put some effort earlier and carry absolute zero risk unless they still hold the token.

There are ‘‘governance coins’’ that are worth in the billions that only serve to govern a protocol. And here, POKT is giving away this value forever for completing some tasks.

You say if my desire is to give the majority control to the whales, this is how to do it. But we all know you yourself are somewhat a whale and you work actively to support the project. Not all whales want to destroy the assets they hold. Also what is wrong with people who have most to lose steering the direction of project? Isn’t that what decentralization is?

The current structure looks more like an elitist meritocracy than a democracy.

6

As stated before it will be easier to discuss if you propose a new system, it will keep the discussion more focused on what you actually want/propose. However I have a coment on this:

I want to say that a token based vote/DAO membership is not enough. I’m no whale and I probably wont be one in the near future, however I beleive that my opinion and vote should count to. It is not a mater of how many POKT you own but how much you undestand about the POKT protocol and its objectives. The amount of POKT that you own is heavily affected by how much FIAT you own and it does not reflects how much you are really risking.

3 Likes
7

Firstly, I challenge your assertion that the DAO has no checks and balances. The validators already have the power to overrule any decisions that the voters make and I am considering making this more explicit with a hard-coded delay on governance transactions and a governance veto that validators can use. Our DAO seeks to ensure that all engaged stakeholders have an equal say on pushing forward proposals that affect them, regardless of their wealth, but as a PoS blockchain the validators hold the ultimate power. This is an example of checks and balances. Whales can’t push through decisions that benefit only them at the expense of stakeholders. But they can veto decisions, pre-approved by stakeholders, that would negatively impact the long-term value of their stake.

Taking a step back, we are indeed considering improvements to the existing system. In fact, some of the critiques you raised here (core team being “gatekeepers”, the general quality of some of the trophies) are well-known and will be addressed. The other critiques about “skin in the game” seem to be ignorant of the checks and balances I described above.

All in all, the current system definitely has room for improvement, which we are working to address, but that’s certainly no excuse to throw out a system with a lot of potential and replace it with a system (simple token voting) that is known to be fundamentally flawed.

2 Likes
9

Then what do you think about DAO voters retaining their right to vote forever? Don’t you think that is too much of a privilege to give to someone forever with no guarantees they maintain stake or interest in the protocol?

10

It is that way for two reasons:

  1. A view that stakeholders who obtain a vote should have their voting rights protected, a) as a defense against permanent takeover by attackers who temporarily gain power, b) to uphold the principle of credible neutrality, by preventing a tyrannical majority from persecuting minority classes of voters
  2. The technology stack as it is currently designed being unable to technically know if someone who obtained their vote continues to be a stakeholder

The first reason needs to be maintained for the sanctity of our governance principles but if a solution emerged that enabled us to remove former stakeholders (in a manner that was credibly neutral) then I’d be open to exploring that as a governance upgrade.

1 Like
11

I could foresee a system in which some very minimal level of of maintenance is required… eg a dev making at least one PR a year (even if not merged); a contributor proposing at least one PIP/PEP/PUP (even if it does not pass); a social media person making at least high-quality post that gets x amount of views or likes (even if none are from current voters) etc. I think that is minimal enough to prevent the issues Jack raises while making sure “dead” persons don’t stay on the roster

12

Isn’t this true of consensus-breaking changes only. A voting majority could pass a PUP that radically changes the landscape, like RTTM or one of the PIP-22 parameters, and I’m not sure the validators have any veto or other power to stop it. Is there a mechanism I am not a ware of, or are you referring to the rule of waiting for 67% of validators to upgrade to a consensus-breaking upgrade before activating?

13

Validators control which blocks are produced. The entity that can change PUP parameters is the DAOOwner (a83172b67b5ffbfcb8acb95acc0fd0466a9d4bc4). If validators are not aligned, then they can coordinate to publish a different block without respect to what the DAO voters wishes or even chose a different DAO owner. It’s not really trivial for validators to do this, or veto right now, but that’s the general idea. Validators ultimately have the final word, unless a social fork happens. I’d hope this would never ever come to fruition but basically here’s the two directions it can go and both can happen:

  • Validators will need to form a new DAO
  • DAO socially forks POKT and spins up a new genesis with different validators
14

Validators technically have the power to block any transaction and all governance changes (parameter updates, treasury payments, protocol upgrades) are enacted through gov transactions.

This has a coordination cost though which is why I’m considering upgrading the protocol to add a gov veto transaction that a 67% majority of validators could use to block any gov transaction.

1 Like
15

Thanks @JackALaing and @poktblade for explaining this. I recommend that any gov change that reduces veto friction be deferred until PUP-14 or similar passes. There is, IMO, way too much centralization risk to make veto easier with only 1000 validators. There are about 4 contingencies (node or pool providers) that realistically could take over 67% of the validator slots at the drop of a dime. Most of them have at least that many nodes sitting at ~60k and for chump change could addstake a hundred pokt or so per node to take the slots and enact a veto to their personal benefit and the detriment of the community

2 Likes
16

All these assumptions are based on the fact that DAO makes the right decisisions. @JackALaing already stated that validators have the power to overrule any decisions voters make.

What you talk about here is censorship by validators being easy, I agree it should be harder. But on the other hand, entrance in the DAO can also be easy for some people, especially those who already have contacts in the DAO ‘‘gatekeeping’’.

If we want something decentralized, we need trust minimization. As much as validator veto should be made sufficiently hard and monitored, DAO membership should also be made sufficiently hard and be monitored.

We know that being a validator requires a certain amount of financial value at stake continuously. But when it comes to DAO, voters only have to get in once and have infinite right to vote without any checks and balances to ensure that they remain reliable and devoted.

This is what I say when I mean trust minimalization. For the veto action, It is objectively confirmable that validators carry stake and financial risk. And they carry these risks constantly or they lose their power to veto.

But in contrast, we have no way of objectively confirming if DAO members also carry risks within the protocol or whether they are incentivized to act in the best interest of the protocol.

In a trust-minimized environment, we should make sure that DAO members and validators are at all times objectively incentivized to act in the best interest of the protocol. Favoring the decisions of DAO members over the veto rights of validators is centralization. DAO members should be held with the same scrutiny as validators’ right to veto. They should have to prove periodically that they are engaged and are incentivized. If the tech stack is not allowing it, then the missions for DAO onboarding should be changed into something more calculable and monitorable instead of relying on outside influences.

Also, I just want to ask, if there is a protocol in place to ban someone from governance if they openly breach or violate their duties as a DAO member?

17

See “Removing Voters” in the Constitution. It is outdated because Aragon Agreements wasn’t released with the functionality that we anticipated. This is one of the parts that will be updated when the next round of constitutional amendments is proposed.

1 Like
18

Thought about this some more, wrote my thoughts in this RFP: Increasing Validator Security through Delegation and Other Ideas · Issue #1477 · pokt-network/pocket-core · GitHub

"Servicers have more at stake than validators in general. It is a bit odd and not in good health that only a small portion of actors in the network gets to be the sole deciders. Delegation allows for a more fair system while also not introducing much overhead to the network. "

1 Like