From mirror by polynya August 9th, 2023

DAOs are meant to be an abbreviation for Decentralized Autonomous Organization. Such an entity would actually necessarily require no governance. Adding in governance makes it definitely not autonomous and arguably not decentralized. Largely because crypto is all about tokens and speculation, DAOs have thus headed down the public companies model, with tokenholders as shareholders. The only real innovation here has been skirting securities laws and thus increasing accessibility to tokenholders. Following the ICO mania, “STOs” were a brief fad - basically public companies adhering to securities laws on blockchain. Needless to say, these went nowhere as they missed the key selling point of DAOs as we know them - evading securities laws. Meanwhile, accessibility for registered securities have improved significantly by Fintech, though still have further room for improvement. It’s certainly possible and likely that some registered securities are floated on public blockchains alongside their traditional rails in the future, but it’s a marginal gain at best.

Before starting this polynya pseudonym, my prior crypto interest was Steem, social networks & governance, and I have spoken at length about how cooperatives were a better model for DAOs in many scenarios. I have also mentioned this idea on Twitter several times, but never wrote a blog post about it. In mid-2016, some of us Steem community members founded a curation DAO called Curie, which is still active on Steem’s fork Hive. This would be make it one of the earliest DAOs that’s still active and sustainable. Curie never had a token, never had public ownership by rando tokenholders, but remained accessible and open, and run democratically by its workers & users.

The key is taking inspiration not from public companies, but rather cooperatives. Now, cooperatives come in many shapes and forms, and so can their DAO counterparts.

The simplest form would be what I’ve previously called Worker DAOs. A lot of smart contracts are simple and straightforward, some may even be immutable, while others require only minor and occasional upgrades/parameter changes. In many of these cases, you don’t necessarily need a huge DAO with many owners. The developers can themselves run the DAO democratically, with a 1-person-1-vote model. New workers can be appointed by the existing ones. Now, of course, this is not going to work for projects with upgradable contracts holding billions of dollars, but particularly in a mature state, you don’t want tokenholders having sway over the same anyway. I expect the endgame for a lot of smart contracts to be immutability, at least the parts of the protocol storing value.

An imperfect hybrid is also possible: Worker DAOs can raise capital by selling profit-sharing rights, but not ownership rights. So, you could still get a token that crypto loves, but it’s not used for governance - just for investors and profit-sharing via protocol fees.

Now, cooperatives just mean a group of people democratically owning the protocol, as opposed to plutocratically by stakeholders/tokenholders. It doesn’t have to be workers - it can be users of the protocol, developers that build on it, other protocols that interact with it, or a combination of the above. It depends on the protocol, where you can fine-tune the ownership structure that’s best suited.

How are cooperatives different from non-profit organizations and democratic governments, then? Of course, it’s the profit motive. The actual innovation of smart contracts is where you can have an immutable contract with no owner and no fees. Where a cooperative-style DAO would differ is there would be a reasonable protocol fee or revenue stream in some way or another. Some protocols certainly behave closer to small governments than cooperatives, so we can definitely have experimentation around such a model. Indeed, Optimism’s Citizen House rings closer to a government than a cooperative, as a contrast to its entirely plutocratic Token House.

Lastly, the public companies with plutocratic ownership model is a tried & tested one that has been successful for centuries. For some DAOs, tokenholder ownership will continue to be an important method, if not the only one (as seen in Optimism). I just don’t think it’s the only one, and many DAOs will be better off with a cooperative model, with a hybrid of revenue-sharing tokens and capital raises if required.

Then there’s being an actual Decentralized Autonomous Organization and not having any governance at all - like I mentioned, that would be an actual innovation over traditional models discussed above. Bitcoin is the closest to achieving this goal, and I hope Ethereum gets there soon too. (As a side note: what about proof-of-work/proof-of-stake? - yes, they are plutocratic in nature, but also a systemic mechanism, rather than an ownership one. The ownership rights remain with full node operators.) Either way, whichever model is chosen, the goal should be minimizing governance. Or not! Some DAOs may actually be better off very actively governed. Which is all I’ll say - there’s a right governance model that exists for every project, depending on the project, and it’s rarely just launch a token and have tokenholders vote on proposals.