Too Slow: Efficiency in DAOs April 15, 2020

The tradeoff between deliberation and efficiency. What can we do to speed up decision-making in DAOs and distributed governance? Following are the notes, as usual generously taken by Lauren.
Link to the video https://d.tube/v/gracer000/QmQeZthzbegsFDKrZjsSTetEoReNipWJ4vMVDb3Ude9RE6

DAO Meeting April 15, 2020

Grace

Decision making takes time, and the more fair you make it, the more time it will take. At DAOStack, you could have a proposal that took 30 days, but it could get boosted. These things all took time, but not too much time, but it was always first proposal in got voted in before there were other proposals. In these DAOs that we have invented, we can start to see how fast the centralized decisions are. They can make decisions much faster. How do we create decentralized organizations that can respond quickly?

Gus

That’s a good topic. We are missing the point about DAOs. We should not only be aiming for decentralization, but to reduce the bureaucracy. Sometimes you need to add bureaucracy to stuff. Organizing DAOs could be more democratized, but at the end of the day, it will add transactional costs.

I have been into Colony for a couple of months, because in my view the guys are taking the right approach.

Heyayat

I love the topic. From my perspective, when we think that we need to have everything. Do we need to have all the features at once?

Thomas

The self-organizing pirates of the Caribbean. When they were in combat they were a dictatorship, in peacetime they were democratic. What are the circumstances when rapidity is necessary? Different styles of leadership in different contexts. Different decisionmaking modalities. “Situational leadership” is the phrase. I can’t remember the author.

Gus

Aragon had different committees with different leaders. They had cells of centralized power being part of a decentralized whole.

Grace

I have been thinking about digitizing this. There are different formats for different contexts. We are good, at humans, for knowing what context we are in. If people are all of sudden talking about something, it is probably something that we need to deal with. If it it about what Kim Kardashian is wearing, it’s not an emergency, but if there is a storm, it needs to be handled. There is this gap between knowing that there is an issue, defining it, and deciding which contextual paths it should take. Which path should it go through?

Lauren

Maybe that’s a role of someone, a watcher who would manually decide if something needed to be slow or fast.

Hedayat

Maybe we need to combine centralized and decentralized systems.

Gus

Time equals money. Sometimes you need more time to make a better decision. Voting should be the last resort. We are not migrating to a new system. Tokens work with allowing people to take on tasks, and starting to create reputation. In Colony, with people gaining power by merit propose some work to do. When someone disagrees, than that person can raise emotion against that specific decision. It’s not about permission, it’s about forgiveness. Something will pass if no one object to it. But someone might put a stake on something if they object to it. Voting is as a last resort.

Thomas

This sounds similar the the “Advice Process” documented by LaLoux in “Reinventing Organizations.” The first thing that people have to do is consult with people who have expertise and will be impacted. At some point they have a recommendation, which they publish a draft. Who am I missing? When it has gone through a consultation process, then you ask for principled objections. Now we are ready to take action, and you can only object if you think it makes matters worse, and you lose rep if you do that too often. That window has already closed. Decentralization of initiative. People engage with each other and harvest both expertise and concerns. This process takes a lot of training, but when it works it works really well. LaLoux mentions several large orgs that are already doing this, like Morningstar. Very nimble, responsive operations.

Grace

I like that, rather than a staking/monetary model, it’s a reputation model. If you are being petty, you should pay in rep and not money. I don’t like the idea that you would have to pay money to object to something. But paying by reputation makes sense. If you are in a minority, you might be unjustifiably voted down.

Gus

You might need the majority to follow you.

Grace

Maybe next week’s talk should be about minority rights. This all sounds business-oriented. Being the majority and being popular will create zero-sum situations.

Gus

Minority: people with low support from the community. They can fork.

Grace:

That’s relevant for a company, but not for a community, or town, etc. You cannot fork if you are a minority group in a city. You can’t just fork off, those people can’t fork without major financial damage to themselves.

Thomas

In any kind of collective, we have to make a decision, and we cannot get to unanimity easily. What if we had to decide, and unanimity is very hard? Collective choice. The situation will always involve a majority. Even though it’s not unanimous, it has a universal impact. One of the advanced in civilization is that when property is taken for common uses, the person whose property is taken must be compensated. This need to acknowledge what benefits most of the people most of the time can really cost people, like when you build a highway built through a minority business district.

Gus

If you don’t like the collective decision, you can take your stuff and go to another community.

Grace

Conviction voting: asserting sentiment over time. I’m not trying to get into methodology.

What if there is an outbreak of lice in kindergarten? Trivial problem, but needs to be dealt with. How you do determine that you have a couple of weeks to solve it, and how you nominate people to solve it.

Martin

Advice process: well-thought through do-ocracy. If you see a need for something, you act, but be excellent to each other. (Martin really likes a do-ocracy). Spiral dynamics. Very strong role of accountability. Do-ocracy. Do it first, and what value is prescribed comes later.

James Waugh

Reputation systems: have to consider all kinds of reputation. We try to boil that down to numbers and scores.

Thomas

I might have someone who walks my dog, and I totally trust that person, but I wouldn’t take investment advice for them. Reputation has certain components to it. We trust people on 3 dimensions:

  1. Do they deliver results? (capability) Does the dog get walked? Does he like you?
  2. Do your words match your deeds? Do you show up on time? (Integrity)
  3. Benevolence: am I in it for me? Win-win, or win-lose?

These three were put forth by Blakey “A Trusted Executive” (see https://www.youtube.com/watch?v=ajXrbaUuSTE)

Lauren wonders if “decentralization” is stupid. Isn’t the problem externalities, like with powerful people having too little accountability? If powerful people were actually responsible for their actions, then centralization wouldn’t be such a problem. There is nothing inherently good with decentralization. The KKK is decentralized.

Martin: Viable systems thinking can help us with this. You have dedicated roles, how to attenuate the internal org to external circumstances. Within your org, you create sufficient complexity. You don’t go into it with a constitution to rule them all. With a slide approach, it might restate in a useful way…it might be worth factoring some expertise in. Parameterization of humans.

James

People change. If you try to define people’s repuation, you are going to lose the full picture. It must be continuously updated. Profiles should really be several dimensions.

Martin

We need to start factoring in complexity. And devolve to humans where humans are good. Be compassionate. Build in ethics.

James

I wanted to share a link to a post I wrote a while ago involving sliders representing values in various dimensions. https://medium.com/@jwaup/everyone-is-multidimensional-2eaf30b1fac9. Excited about using color for reputation.