DAO Design Elephants

DAO Design Elephants  v220422

A lot been said already in DAO nerd circles about DAO design patterns, and this topic becomes less and less nerdy.

There seem to be a few elephants in the DAO room that everyone encounters at some point. They might appear like conversational deadends, or even as limitaions of DAOs themselves.

Even though, I believe that everyone is better off finding their own path to those elephants and meeting them in person, I want to try to articulate some of them here. This is not yet another attempt to formalize DAO design patterns, it's just an invite to share experiences. In the end there''s a link to this draft in Cryptopad, and your comments are much appreciated.

If you have been in a few DAO design conversations before you'll see what I'm talking about. I'd like to articulate it better, and shorten this text as much as possible, this is the first draft.

  1. Identity. Every conversation about identities, a basic building block we have to create persmission systems, ends up in the conversation about sybil resistance. Any (theoretically) perfect sybil protection is deeply dystopian.

  2. Who can make a proposal? This refers to 2nd layer governance DAOs where sentient beings are voting on things, not the algorithmic ones. This is a rabbit hole, I write more about this in the section 6. But I think it deserves it's own section.

  3. Granularity of decision making. If you try to do everything by concensus the system becomes impossible to navigate. I wouldn't call this "death by bikeshedding" because the decisions can be important. When do you need a vote? How do you decide when the new vote is needed? (This one triggers some kind of fractal vision ocd in me)

  4. How to price work in DAO own tokens? What is a fair price to pay for work? In the Dark Ages where the CEXes ruled Cryptoland, bounty hunters would accept a shitcoin that might go up (or very likely down, unless the issuer provides a wet sopping liquidity) when it hits an exchange. "Fair" is a slippery term. We could start comparing prices in the normal job market, we could try to be ethical. Thing is, there's no right answer to it, and by the way, even though you can dump mopst tokens easily on a DEX nowadays, you never really know what gas fee you're going to pay and what

  5. How do you ensure that the work that has been done is done well? Someone has to check. Lets create a quality assurance role! How do you make sure that this quality assurance is of good quality? This someone who checks is not malicious, and checks properly. And it's just crypto-police all the way down. It is a deeper philosophical discussion, that deosn't have an answer outside context. It deserves a post of it's own, a discussion and a reading list. I leave it for now, in my personal opinion "policemen all the way down" is not the DAO world I want. Next point is closely related, I might combine them into one.

  6. Fear of being not permissionless enough vs fear of freeriders. A struggle to start as decentralized as possible. When setting up a DAO a lot of people would want to be decentralized from the start. But how decentralized? Nothing is really permisionless. Permissionless systems we describe are systems where permissions are granted if you follow a rule. On the algorithmic layer it means, for example, that a free transaction cannot be accepted by the network. On the human messy layer it means that you still have a multisig, and not just posting a private key of a funded address out ion social media waiting for someone to take it and dufuly complete the job you wanted them to complete.

  7. Trying to apply project thinking to emergent systems. Cryptoeconomics thrive on the idea of a rational actor. One of the images of a DAO can resemble to a Rube Glodberg machine behind a clear glass. Perfectly auditable, automated. While smart contracts can certainly provide some amazing degree of cryptographic verifiability, the moment the humans are involved everything becomes messy. You can create beautiful processes, but no matter how beautiful they are, they don't magically turn into practices. Groups of humans thrive on rituals & ceremonies.

That's it for now.

I first wanted to call those DAO Design Vices. Like 7 vices or 7 deadly sins. Then I thought I could call them curses, because they're like magic spells that send conversation in circles. But I'll just call them elephants, a little like elephants in the room, although nobody wants to avoid them in the conversation, they're just standing there.

What would be cool, is to create some art work associated with those, give them personalities, characters, like, I mentioned, vices or virtues, or any magical creatures. Have I missed some?

Just to conclude this post, that is tricky to finish, even the first version. I don't think these things I mentioned are real limitations. They only exist in the framework of mind we are thinking from. They only exist in the current norm of the world, they're not some real physical limitations. Beyond our current frame of thinking those things will not send conversations in circles at all.

This is the view link to this text. https://cryptpad.fr/pad/#/2/pad/view/oIEYEckAiFuEfZLMcyXZ67UWV-tLXriX3rohZmretlw/ please leave a comment.

Other vices and deadends of the mind Here are some dreams and things that frequently occur when any group comes together to do something. This is not really specific to DAO design. I thought I'd list them here too.

  1. A dream of perfect documentation. All documentation gets outdated, and in our sphere even faster than anywhere. There's a special palce in hell for outdated documentation, it's really can be worse than no documentation at all, because it is misleading and noisy.

  2. Choosing the right tools. For organizing, for documentation, for communication. Oh let's choose that perfect tool that everybody is okay with, this noce central place we come to and everything becomes clear and easy and easy to find and to follow. This doesn't exist. It will always be a mess, even if you stick to some seemingly clean thing like google apps.

  3. "If you pay them they will come." Somebody will come, but real longterm community members I'm not sure. Giveaways is a legit strategy & an age old trick to get people through the door. But it's only complimentary to what gives people meaning. With more capital than skill, and even in bearmarkets, cryptoland is full of bounties and airdrops. Rewarding your contributors is incredibly important, not just for the contributors but as a fundamental primitive of token distribution and DAO design. But, if you just throw tokens at random swarms of cheap labour it can backfire. Paying people is complimentary to the culture a community actively creates. If people are performing tasks that make no sense to them, or not fun, they will always go somewhere where they are paid more. And in cryptoland someone will always pay them more.

I stop at 10 in total.

Luckily more and more people are actually trying to set something up in practice at the moment. We don't reach these deadends of our organisation design thinking theoretically, we reach it practically. And that's the only way to break through this threashold.

Million times we will repeat this dance, dancing in circles around these concepts, trying to find a practice, million times we will recreate old rituals and ditch them. One person understanding something, like Cassandra, doesn't really change anything, shared experience does. It changes everything.

Concepts and things that helped me to gather my thoughts: