Evolving Algorand Governance

I think Foundation and Saith are doing a good job - they elaborated one idea and are seeking community for opinions. We can’t build the best governance system over a night. It requires good engineering team and evolution.
That being said, I don’t agree with the current proposal but I trust it will evolve in the right one.

2 Likes

I would be very surprised if this (daily average) can’t be gamed in which case it’s not a safe solution (you also need to trust the source of this daily average, which is another issue). I believe that raising TVL by centralizing the system is not the right way to do things.

Also who’s responsible when things go wrong? Let’s imagine this happens: selected group of people say “dapp X is part of governance”, people then put money in X, someone hacks X and steals people’s money → what do we do now? We know there’s no way to guarantee that a software has no bugs, even if it’s checked by 5 different companies and dapp is formally verified. This is going to happen sooner or later and people will not only lose faith in the hacked dapp but in this case also in the algorand team for including such a project in the governance.

I believe that if we want to raise TVL we should do it in a proper way, not by centralizing the system and making the main algorand team a target for other people’s mistakes. We have the team that came up with the best pos consensus and with state proofs, so i believe they’re more than capable of coming up with a decentralized solution where algo stakers decide which dapps to add/remove from the governance (if we decide to go that way). To me a much better quick solution to raise TVL is to lock the coins that are staked and then over time figure out a decentralized solution to incentivize people to use defi. Let’s not forget that decentralization is the main reason why blockchain is a powerful tool and that decentralization is what enables us to make the world a better place.

3 Likes

This language isn’t appropriate and explicitly attacking individuals is not the way to solve this. I believe the issue may be the incentives and the strategy that the Foundation has which right now, with my limited understanding, seem to be “we have to maximize TVL”. This is why slips like this one can pass through. This proposal maximizes for a metric that’s been relevant for a year or two while introducing some very shady (at best) additional democratic votes with a pinky promise that the added third parties will be fair and can hence be trusted. The side-effect of this change is that serious investors that are in it for the long term will see this as a panic move to become/stay relevant because the described change has no solid grounds. Rather than introducing trusted third parties, we need to keep the democracy fair and verifiable by everyone. Regardless whether the party would be the Algorand Foundation, defilama or Jesus himself - not even he is trusted by everyone. Let’s try to keep the democracy clean and simple and if we must add something, let’s at least do so through research.

3 Likes

The way I see it, governance (even of a platform/ecosystem) goes hand in hand with politics (of the platform/ecosystem). This is not to say that we should bring into the discussion the world politics. However, we do not have to reinvent the wheel, and I believe it helps to have analogies from daily life to help us see this.

I think it is quite clear (at least for the Foundation) what the Governance is supposed to achieve. See this from the recent Community All Hands event. I completely agree with this vision. Here are my thoughts on a bit more concrete goals for Governance.

The road to that vision is in the first step the implementation of a decentralized governance system, which is what we are (or should be) addressing. In my personal opinion, the most decentralized, transparent and fair way is direct participation, which are also the core values represented by PPoS. That’s why I suggested to implement Algorand Governance as referendums.

However, the current discussion seems now to focus on even the right to vote and the voting power, which has completely overshadowed the discussion on Governance implementation.

What I find most baffling by all this discussion, is that the right to vote and the voting power was already decided at the initial network-wide referendum on the governance program. In the details of that referendum “Decentralizing Algorand Governance”, accessible e.g. here, it is clearly stated at page 4:

That referendum passed with more than 90% of online stake.

6 Likes

It seems to me my message might not have been clear enough.

The only asset that I consider that should have the right to vote on the future of Algorand, is Algorand (i.e. ALGO) alone. No other assets. Allow me to try to elaborate again on why.

ALGO holders have a stake in Algorand, thus are inherently incentivized to do what is in the best, long-term interest of Algorand. Any other assets that come to Algorand, require only a minimal ALGO stake. Therefore, they lack such an inherent incentive. Hence, if they were given voting power, others would need to blindly trust them to do what is good for Algorand.

However, the only premise that should be required from the ALGO holders to have is that each other user will do what is in their own best interest (i.e. that they will not be self-destructive). This is the basic assumption of PoS that guarantees the security of our funds. Trusting in nothing else should be required from us, regardless of how good we think non-ALGO assets might be for the ecosystem.

Further, the idea of 1 ALGO = 1 vote, was also already decided at the initial network-wide referendum on the governance program. In the details of that referendum “Decentralizing Algorand Governance”, accessible e.g. here, on page 4:

That referendum passed with more than 90% of online stake.

If that were not enough already, M1 is suggesting the assets to be given voting power to be completely liquid, further minimizing the required commitment towards Algorand.

7 Likes

I agree that having long governance periods is slow and can be inefficient. However, we do need to have them. We cannot have a completely liquid governance because voters should be required to have a commitment to live with the consequences of their votes.

Further, if you had too many voting periods, many would not want to participate due to it being too much work. While some might say that is a good thing because only invested and informed governors would remain, I think we should rather try to make it as accessible to as many as possible. This is the only way of making it “very” decentralized. With this in mind, I think we should try to keeping it simple and inclusive.

Not having governance periods is not an issue in Cosmos because they use bonded PoS.
Having too many voting periods I think is already being seen as an issue in Cosmos. Some chains have really frequent votings, and also on very specific issues on the protocol level. That is why a lot of people simply automatically vote yes. Rarely I see people admit they are not informed enough or interested enough to abstain from the vote.

Therefore, I currently think quarterly governance is a good compromise for both of these aspects, and the updated suggestion as its implementation possibility.

I would not set this as a requirement because it could turn out as a single point of failure. But it could be a recommendation. And in practice, if a proposal would not be discussed first, it would most likely not gather enough support to even come up to a vote.

2 Likes

Very frequent voting may come with big downsides. Fast decisions lead to poor choices and each implementation adds new elements to the system which make the system a spaghetti in the long run.

2 Likes

I think there needs to be a serious re-evaluation of everyone’s role at the Foundation who was involved in finalizing measure #1 option A. There’s plenty of discussions around how potentially detrimental the option itself could be, but not enough concern about how it could even pass deliberations to begin with. It’s not something we should overlook as simply the worse of the two options. It shows a very worrisome level of negligence and poor thought process. This is a cut-throat space in which we should not settle for anything but the best, and therefore we should not allow the same people to continue having an influence at the Foundation, especially in areas which they clearly do not have a firm understanding of.

1 Like

@tebok That is an excellent counterpoint you make.

Just to make sure I understand it correctly, as a concrete example, you would argue e.g. that current ALGO holders might not want to decrease opt-in costs for an asset at a point where the price of ALGO is high, because new users would be forced to buy a larger ALGO stake in order to be able to participate, driving the price of ALGO even further up?

I do see something like this could indeed happen, but I think such a situation would naturally resolve itself. Because if existing ALGO holders were to start doing this, new users/projects would seek other platforms to rather build on/use, decreasing demand for Algorand and forcing it to again become attractive for users.

This is actually similar to what is currently happening to Ethereum (there you have the miners resisting change instead of existing holders). But as you see, competitors are emerging that provide better solutions (e.g. Algorand), forcing Ethereum to upgrade. You could argue that Ethereum has had issues already for many years and no one has managed to overthrow it yet. True, however, I imagine that in the future it would be much easier for a competitor to raise because the switching between platforms will be made much easier through development of blockchain interoperability. This interoperability will drive competitiveness, preventing what you mentioned in the first place. That would be a problem only if one (i.e. Algorand) were to dominate the crypto space. But as Silvio frequently says, I do not believe one blockchain will dominate the space, but we will live in a multichain world.

1 Like

@omniwarp, you mentioned a research paper, we actually came out with one around two months ago

There we mentioned allowing governors committing algos in Defi for a term (via vaults, liquid governance or possible DEX commitment) to make voting proposals, and receive rewards suitable for their term commitment in an economic activity. Before and after that there was a large debate till this proposal, but the background motivation for the involvement of DeFi in governance, given its economic relevance, should be strategic. With all the improvements we can bring in the process.

1 Like

Thank you for bringing attention to this article.
@Massimo, in the article you wrote:

Could you elaborate on the reasons why you now see a need to change these fundamentals by allowing other assets to participate in our common resources and our governance?

Such a move seems to me to be driven purely based on short-term economic speculations, and by far not in the best interested of Algorand since it can even be extremely damaging to it - as I have argued and demonstrated.

2 Likes

I think the Foundation ought to have a team whose sole responsibility would be the Governance.

There have been too many issues with Governance so far - e.g. incomplete proposals both during G2 and G3, not announcing the topics to be voted on before the start of the governance period, changing the topics as in G2, and now last-minute changes to a proposal before the vote. Based on just looking at the number of responses in this forum from the Foundation compared to the number of well thought of comments/concerns raised by the community, whoever is currently responsible for it at the Foundation seems to be overwhelmed.

Massimo is chief economist, Shai a research fellow on cryptography, and I am sure their talents are utilized on their respective fields.

Hence, I think a dedicated team (of the highest integrity) could improve the situation drastically. Its top priority ought to be the transition of the Governance from the Foundation to the community, and its continuing mission later to ensure the execution of the Governance decisions.

4 Likes

I had assumed that the Foundation already had a team that was dedicated to governance, but if they don’t, you’re absolutely correct that they need one. The xGov program will undoubtedly take a significant amount of time to fully implement, and the Foundation will still need to take a leading role after it is in place. I’d maybe even apply if they opened up a few governance team positions.

At a high level my main point was that if the current Algo holders were primarily investors and market speculators, they would be financially incentivized to try to drive up the price of Algo (in hopes of cashing out at some point), and such incentives are not automatically the same as what is in the best long-term interest of Algorand. (If by long-term interest we think of Algorand’s wider adoption for the purposes it was initially created for.)

I was also trying to show through the analogy that there may be other groups of (potential) Algo users who may not be large Algo holders at the moment, but who would have significant real world use for a widely adopted and technologically advanced crypto like Algorand, and who therefore are incentivized to act in Algorand’s best long-term interests (rather than treating it as a speculative asset where the underlying technology is secondary to the market price of the asset).

The conclusion therefore was, that relying on the market forces of current algo holders via “1 algo = 1 vote” may not actually be optimal, and can lead to governance decisions that sacrifice Algorand’s long-term interests in favor of decisions that increase it’s market value in the short term.

Unfortunately I don’t have any better solution to suggest either. At least none that wouldn’t make governance political (where people rather than algos vote), and I’ve come to accept that very few people in the crypto world want to go to that direction. Therefore “1 algo = 1 vote” may just be the most “democratic” form of voting that there is. Though I would also hope that before “1 algo = 1 vote” fully takes over, the ownership of Algo was much less concentrated than it is now, and that most of the Algo holders were also using it for something other than just for market speculation.

My general issue with both proposals is that they seem to overly reward being rich. These proposals seem to be rewarding the richest with responsibility, instead the other way around.

I’m all for a rewards model that compensates for the level of responsibility taken and fulfilled. To the degree that there are different levels of responsibility (DeFi, Relay Nodes, Passive holders), there could easily be different reward tiers.

It feels like this proposals allow for a wealthy person uncommitted to the ecosystem to enter and co-opt the xGov tier., make self rewarding proposals, and then get extra voting seats to approve them.

Is this incorrect?

2 Likes

iirc from the xGov proposal was that xGov cant vote on the finished proposal. so they could theoretically put forth something that they like and then pump it up so that it goes to a vote, but then they cant vote on it on governance.

I agree, and I think it’s paramount that the transition to community governance is done in the most careful and responsible way possible. In this fast-moving space, you cannot make the most informed decisions on a particular topic if your efforts are split elsewhere. Like you said, it’s evident that something is not going as smoothly as we’d hope for at the Foundation in regard to governance. Just voting away the questionable measures won’t fix that underlying issue, and will likely continue down the path of disorganization and head-scratching proposals (and decisions) in the future.

One possible solution is to lock the governed algo for some time (longer than the gov period to avoid cheating short-term, eg. if unlock happens exactly when the gov period finishes the whales could coordinate a last minute vote-swing and then dump algo because they expect the result to drop the price). However even when algos are locked there should be an option to participate in the next gov period (just extend the lock time in this case). The longer they’re locked the more it incentivizes long-term decision making.

2 Likes

One additional advantage of locking staked algos is that exchanges can’t vote, at least not with the majority of their coins unless they want to risk running out of algo and hurting their reputation. Which in turn also means regular stakers get more rewards

Thanks for the link to the blog post. I apologize for the late reply. What I had in mind is an actual research paper with math showing why this works just as well if not better. For instance, let me point out some logical conflicts I see from the option A. I will compare Algorand consensus and governance. I’m aware the two play different roles, but they share the same environment. Let’s first agree, that both consensus and governance operate on the same assumption “the majority of money is in honest hands”. They also both recognize 1 ALGO as a citizen with a power of 1 vote. This means their definition of an Algorand citizen with a vote is the same.

From what I understand, the Foundation claims to have found a better definition of a citizen, meaning more inclusive and thus more decentralized, by recognizing some other (defi) assets as citizens as well. I agree with the part of the blog post that suggests that the more citizens we have, the better. However, if option A is in fact a better citizen definition, since both the consensus and governance operate under a definition of a citizen, by implication it would mean we should use the “citizen definition” described in option A on the consensus level as well - because it improves inclusion and decentralization. This would mean that defi participants could participate in the consensus with their own VRFs weighted by their TVL or whatever way the votes are evaluated. I’m fairly sure this won’t happen at the consensus level and my question to the Foundation is, why not?

My best guess would be because option A’s new way of labeling citizens worthy of voting in fact comes with new potential security holes and trust assumption as was already discussed above in depth which would be extremely dangerous for the chain survival if we applied these definitions to the consensus. This then begs the question… are we saying we can tolerate less security in governance to resolve some other governance related issues e.g. clashing of defi and governance as described in the blog post?

For what it’s worth, I actually think this is an honest attempt at improving governance, but it seems to me it slightly lacks solid grounds. If the grounds were in fact very solid, the same logic of who is a citizen of Algorand should be considered at the consensus.

2 Likes