UPDATE 23/08/23 : almost a year after pausing research on the idea, it seems like it was nearly spot on. It is time to reactivate that debate. Today we have stride in the ICS and it brought some significant similarities to our model. The first of them is the liquid staking of course but most importantly, it brought validators with no consensus risks… Basically this is almost equivalent to what we highlighted in the following document. It creates a potential for a political class to receive token delegations and take a fee without risking any slashing. It’s exactly similar to our prediction of separation of security and voting. We think the market will mature over time and stride may take a significant share of the cosmos politics in the future. This is also where our prediction for councils will likely emerge. This is an interesting time to be in Cosmos !
UPDATE 21/12/22 : After a few weeks in that forum, we have been disgusted but the low level quality of the discussion. Despite a few people who tried their best to debate on the subject, most of the participation is trashing non-sense. Anyway we wanted to let you know that our first experience with cosmos forum is disappointing to say the least. We decided to migrate to another ecosystem. If we have one advice left to share as a positive feedback for those who care about making this forum a better place for the following initiatives : we expected genuine talks with smart people, willing to compromise & help us improve our ideas; That’s what this place should be. It is far from what we have experienced, sadly.
Introduction:
I’m Phil, co-founder of the Govmos initiative (more on that coming later in this forum but that’s a topic for another day). with this thread I’d like to introduce the research that I made recently on the complex topic that is governance. The Cosmos ATOM 2.0 proposition introduced a very ambitious project around Treasuries & Governing Councils. In this paper I wanted to deep dive on the initial idea and see if we could build a viable socio-economic model out of these two new tools.
(I’ve been exporting images from the original PDF file, you might want to zoom on these to read properly) if you want the original PDF version (with text selection) here it is : Governance Structure Proposal.pdf - Google Drive
Conclusion:
Thanks for reading, hope to get constructive feedback from the community. I’ll try to answer your comments as quickly as possible and update as these research topics evolves.
I have rarely been able to read something so unclear. Obviously the understanding of liquid staking is not good by the authors. A crazy solution for a non-existent problem. Thanks for nothing. Could you elaborate on the problem you think you see and thus solve?
How much would the allocator cost/month? and in total? how much would the allocator return to atom delegators?
far as i can tell, 2.0 incorporates a parasitic layer of leverage in LSDs and parades the growing number of supporting chains required to attempt correcting the incongruent dependencies that emerge as ecosystem growth. dependence on other sovereign chains increases complexity and risk of exploit and bleeds value from Atom to sovereign chains.
if the only reason ICS chains use ICS is because they are heavily subsidized by Atom delegators, it is not a sustainable venture.
you could fix a majority of governance related objections to 2.0 if you make it so no validators can vote with delegator votes.
seems like validators have to much power over the direction of chains. this creates perverse incentives and provokes political melodrama.
validators get rewarded for signing txns, not mess around with governance and propose a restructuring of ATOMs monetary policy to increase txn volume at the cost of efficiency, security, and token holders.
Its only a rough estimate here, but depending on the native chain the treasury can either be a module in the main blockchain, or a “sidechain” type of deployment using shared security. In the hub’s case I think it’s safe to say we’ll go for a sidechain to keep the hub as minimalist as possible.
On that case the cost of running that chain is the same as any chain it’s the cost of running the validation process which is arguably negligeable considering the hub’s marketcap. I don’t have the real numbers here but a validator instance on a cloud can cost between 500-2500$ multiplied by the nb of validators.
However, on the return side the value is easy to understand. We seriously increase the governance decentralization (as explained in this paper) while also reducing the cost burden of delegators. In that system the governance organs are funded by their referring cohort (validators pay for validator councils, etc…). This seriously increases the accountability of the councils; They will have to bring value to their cohort otherwise they’ll be defunded and replaced by more productive alternatives. Imho this system is the best I’ve seen so far to incentivize cost optimization of governance. The only additional cost can come from deploying a new chain on shared security instead of a module ran on the mainchain. But on the hub’s case, the minimalist approach could definitely justify the cost. It’s up to the community to debate on that front. I’m personally OK with both options
this is a reductive vision imho. In the paper I insist on the fact that governance has to be an opt-in system. That’s the case for every cohort willing to participate. Validators definitely might want to participate and they’ll have the rights to. Some will be just “tx processors” but some may participate in governance as well. On a lobbying vision eventually to defend their business margins or in a broader sense. Remember that their business depend on the blocks being demanded, if the chain goes less active they’ll make less revenues. SO… they definitely have a word to say in that governance structure.
Where I agree with you is in the fact that today they are forced to do governance because otherwise we wouldn’t have quorums to pass votes. My model offer partial solution to this, but the true change has to come from the increase in participation of all cohorts. That’s something we’ll dig deeper into when I’ll disclose some of the things Govmos will offer for this specific problem
how many delegator votes are validators casting vs delegators casting themselves?
nothing stops validators from self delegating and participating, using others votes to direct the chain to profit themselves isnt a right validators should be entitled to. an opt-in system only centralizes the chain around aggregate voting power of individual validators.
there is likely a workable paradigm for validator controlled votes that would satisfy quorum and reduce the conflict of interest involved in validators controlling the direction of the chains with other people’s votes.
I have seen validator voting power = the square root of delegations +self-delegation suggested. that seems agreeable to me, but if a quorum isnt possible with that adjustment, perhaps there is a better calculation.
The cosmos should work toward efficiency, utility, immutability…not validator profitability at the expense of those qualities.
I have only one question at this moment in time. When it comes to ‘Delegator Councils’, you mention that these are created for those delegators that don’t have time to participate in governance and decision making essentially and that they should give their vote to a council which they sort of ‘align’ with.
Currently, people delegate (or should delegate) to Validators that they align with or that they like. Validator governance participation has been a hot topic recently and I’ve seen a lot of validators up their game when it comes to this.
So my question is, what is the difference between them and why is there a need to create another layer for this specific council?
It’s a very good point that you have here. I personally think that having delegators council will be the most difficult balance to strike. Still I think in the longer run they have to be there to compensate the overall extended political role the validators will have to endure otherwise; Some validators do play that political role quite well today. But the truth is that as the ecosystem scales I think it’s going to become increasingly harder for validators to do both delegator’s councils & validators council roles.
It’s also important to note that inevitably delegators & validators interests will differ at some point. so if there’s no one defending pure delegators interest in the future that might be a problem imho.
But I do get your point, delegator’s council don’t seem like an immediate necessity. And tbh I agree with you on that. But the key point of my thesis is to see Delegators as pure economic entities, put their capital at risk and look for growth. That’s what they should defend in gouvernance; take their tokens at risks to allocate the capital the best possible way to bring growth. Validators instead should defend uptime, technical viability of updates & sustainable money flow for their job.
In my system, people should choose validators because they do that well. They should also choose delegators council members meant to seek growth for the chain. Pick those they think can create more $ tomorrow than the one they have today. Two totally different roles imho
This sounds a lot more like a patch than a solution to the real problem; In my paper I tried to avoid that and identify simple solution that don’t require too much hassle. Separating powers by identifying which major forces should compete. I came down to that simple solution of 4 cohorts : delegators, validators, app developpers & team+foundation.
We we’ll have these 4 forces participating in governance, there will be less chances to see one cohort controlling the system.
Hello, I’m a simple investor in this ecosystem, and actually a law student.
Governance is a subject that I love pretty much in general. Maybe I have wrong but what you trying to do with the separation between Delegators councils and validator councils seems like what we can see in the congress in the USA with the house of representative and the senate is that true? It smells like you try to build a new state and I am for this initiative.
Like I said I am just a simple investor, I love what we are trying to build in cosmos, but I can’t spend all my time to understand how to grow this ecosystem and it is not my job, I am completely for creating different councils in competition who try to do the best for the interest of the majority that’s how we can see growth coming into the cosmos in my opinion.
( Sorry for the mistakes of language or understanding, this is my first participation thank you for reading me ^^ )
It’s not exactly copying the senate vs house separation. But the key takeway would be that governing blockchain correctly mean that every participant should have it’s own set of representatives in the democratic debate. If we stick to the US legislative structure then our councils would be like the cocus in the house. Because they all defend their own interest (validators defend the tech feasibility & their fee model, delegators defend the economic viability, foundation defend the long-term coherence, apps defend the direction they want in term of development & features). Not having them represented in the debate would mean that one party can set the direction on its own. Which is exactly what the intial “governing council” proposal offered. One council to decide how we should structure governance… that’s the CCP. If we want to make it a democracy then what they proposed in atom2 would be the foundation council, and there would be 3 others cohorts to create (each one with multiple candidates). On top of that the foundation council should be funded by the foundation, not the whole community’s reserve. The paper explains that perfectly. The funding model ensure accountability
I think I know what liquid staking can do. It surely doesn’t allow to transport votes atm but there’s a lot brewing in that direction at quicksilver for exemple. Liquid staking modules will come early next year on the sdk and therefore there are still a lot of things to be added. Splitting financial delegation (destined to validators collateral purpose for security) and voting delegation (destined to cast vote on behalf of the user if he doesn’t) can easily be achieved within the LSM.
That was actually one the main takeaways on that paper. We need liquid staking to do that otherwise we’re stuck with the same old problem (the one I’m trying to solve), validators have a double role and this will likely not be suitable at larget scales. Validators will have hundreds of chains to maintain technically… Asking them to run the politics on all those chains is just not serious. We need governance specific bodies, not just for decentralization, but because validators simply won’t be able to do the whole thing forever. A few may still want to do governance, they may create validator councils.
Anyway I have to say this kind of comment is quite irritative, would be better to be a bit more gentle It took weeks to make that paper, and there’s obviously a lot more depth than what’s been presented here. We just wanted to taste the temperature on the overall architecture first before diving in the details if necessary.
it’s a research on an economically viable structure for DAO governance… It was meant to explore a way to use the proposed tech from ATOM 2,0 paper (treasury modules & governing councils) but in a more extended way than having one treasury and one council. We offered depth and balance of forces. I’d love to run calculations but tbh I don’t understand how they would fit in here. Maybe you could tell me which calculation you would like to see and I’ll run them.
So we agree validator governance is a problem…
I think the problem is that validators have to much power & dictate the direction of the chain by voting with other people’s votes for their own benefit and at the expense of delegators/investors. and the best/most simple way to deal with it is reduce validator voting power.
reducing validator voting power= square root of delegations + self-stake (or similar to account for quorum) solves the majority of governance issues without needing all the complex and messy cohort calculations and such.
validators are rewarded for signing transactions, not playing politics. validator incentives, when voting in governance, are not aligned with delegators. validators overweight political opinions use other people’s votes without requiring any skin in the game & results in a validator centric ecosystem that is un-investable
can you calculate how much less hassle it would be to reduce validator voting power= square root of delegations + self-stake instead of implementing an elaborate balancing act of “forces” you are actively defining and trying to coherently piece together?
Your solution is indeed much simpler; We’ve also seen similar initiative like validator’s vote weight cap (in tori i think). These are patches at best, and the problem stays the same. You just shake the soup and install restrictions, but il will be overcomed. Big validators will just run multiple instances if they want to shape governance. Just like imposing caps on airdrops… it’s not solving anything, people just create multiple wallets (solving airdrop is acually also accounted in my solution).
On the square root + self delegate, well it will just do the same here… validators who would like to influence governance would just self delegate a lot and takeover large portions of vote power.
Actually I also think the problem is more complex than just validators having too much power. It’s a problem of any of the 4 cohorts being part of the network trying to control goernance. Today the risk is around validators, tomorrow that may well be around the foundation if we have a single council (like ATOM 2,0). See where I’m going ?
how are they patches at best? the problem is that validators have to much voting power that it skews governance toward the benefit of validators at the expense of delegators…
if validators (or anyone else) want to buy a large enough delegation to govern the chain that is fine.
What good will multiple instances of a validator or wallets do to shape governance that is similarly unfair relative the current validator centric paradigm we currently see?
buying their voting power is preferable to the current system where governance is surrendered to validators and directed by validator profitability rather than chain sustainability, security, etc.
actually I have no idea what you mean or where you are going. ATOM 2.0 is a validator’s wet dream. more txns to sign the more money made. the conflicts seem to stem from the validators voting with other peoples votes by default, which means the foundation, or anyone else, needs to onboard validators to get governance passed. validators are holding the cosmos hostage through unfair vote weighting.
Blockquote actually I have no idea what you mean or where you are going. ATOM 2.0 is a validator’s wet dream. more txns to sign the more money made. the conflicts seem to stem from the validators voting with other peoples votes by default, which means the foundation, or anyone else, needs to onboard validators to get governance passed. validators are holding the cosmos hostage through unfair vote weighting.
i’m not that pessimistic but I do get your concern on that one. I think validators are not evil, they just serve their own interest indeed. Just like my models says actually. That’s also the case with a foundation council, that would serve the foundations intersts only. Same fore delegators who would just push to increase the token value one way or the other. And app developpers on the chain would do anything to use the comm pool to fund cool stuff of them all the time…
That’s why our thesis is that none of these 4 actors should lead the governance, they must compete against each other to strike the best possible balance. To us, the goal is not to restrict validator’s governance power, it’s to make sure they have opposing powers to do it naturally if they try to abuse. But it’s also important to understand that validators would also be essential to limit any abuse from any of the other 3 cohorts. It’s a multifactorial defence system that should balance on its own
Validators start off with to much power. making their voting power flat self-stake would make everyone equal. I am suggesting √delegations + self-stake because validators are more knowledgeable than an average holder, but not so much that it creates a huge disparity like it does now.
why make a foundation council when there is already the ICF?
so long as validators vote with aggregate delegations none of this matters though, whatever councils decide on will still have to ask/convince validators to get permission to pass it. im not sure such a permissioned environment would attract builders let alone investors.
It took weeks to make that paper, and there’s obviously a lot more depth than what’s been presented here. We just wanted to taste the temperature on the overall architecture first before diving in the details if necessary.