Decentralization is often discussed and debated amongst DAO builders and operators. And rightfully so – decentralization is a fundamental tenet of web3 – so it’s important we’re all working off shared terminology and frames of reference of what it means to be decentralized. Less often discussed, though, is the “A” in DAO. But maybe it’s time we start looking at it with more rigor and care.
This piece is a provocation to explore what it means for DAOs and, importantly, their constituent members, to be autonomous. To help frame this discussion, let’s first look at progressive decentralization.
In “Progressive Decentralization: A Playbook for Building Crypto Applications,” Jesse Walden describes progressive decentralization as, “a process in which founding teams relinquish control by degrees, over time”. It is a useful construct for core teams to grapple with the tension between finding product/market fit and community ownership.
Walden maps this process to a three-stage product lifecycle for a blockchain-based crypto application. In the early stage, the core team makes all decisions on what features to build and how to find product/market fit. In the growth stage, the core team invites community participation in product development but final product calls are mostly left up to the core team. In the later stage, the core team steps away from the product and hands the reins to the community.
Throughout the lifecycle, community participation is coordinated through incentives such as bounties and grants, and airdrops in which tokens conferring voting rights are distributed to community members.
Through this lens, token distribution positively correlates with community ownership. As tokens are increasingly distributed to the community, the community (i.e. token holders) has more influence over the decisions about the application. This “exit to community” is “the moment a smart contract is triggered to mint and distribute tokens.”
This moment marks the point at which the core team relinquishes majority ownership of the application, hence ceding control and decision making power. At this point the project and its governance is “sufficiently decentralized” and “independent of the efforts of a single entity that might have asymmetric information.”
This simplified description of Walden’s conception of progressive decentralization should serve as fruitful fodder for exploring progressive autonomization.
If progressive decentralization is about the “D” in DAOs – dealing with tensions around product development, product/market fit, governance, and community-ownership – then progressive autonomization should obviously seek to deal with tensions around the “A” in DAOs.
But what does this actually mean?
More than anything, here, “progressive autonomization” is intended to spark critical thinking, discussion, and debate. This provocation aims to elicit richer, more nuanced, and shared understandings of what autonomous means in a DAO context. The hope is this will surface tensions surrounding an implied ambiguity around “A” and, in turn, yield valuable, actionable playbooks for DAO builders to progressively autonomize.
Below are some prompting questions I’m most interested in exploring.
Like in precise and practical terms. To me, our answers to this question should be informed by how we understand the relationship between autonomy and automation.
As I dove into this piece I became increasingly aware of an implicit assumption I held - that autonomous-ness mostly deals with automation. As Vitalik describes, “DAOs == automation at the center, humans at the edges” (see Vitalik Buterin’s “DAOs, DACs, DAs and More: An Incomplete Terminology Guide”). But I think automation only answers part of the question of what it means to be autonomous in a DAO context because it has to do with the code side of DAOs.
I’m coming around to a richer understanding that autonomous-ness in DAOs is part automation and part autonomy. While automation deals mostly with code and processes, autonomy deals mostly with people and governance. Maybe.
This begs another question…
We know that DAOs are internet-native organizations that run on blockchains, have internal capital, with automation at the center and humans at the edges (see the aforementioned Vitalik post) . But what does this actually look like?
I’d love to see a graphic “A Relational Map of the Autonomous Components in DAOs” or something like that.
It would help to answer a related question I have around, what similarities and differences of being autonomous at the code layer (i.e. smart contracts that automatically execute on-chain upon specific conditions being met) and autonomous at the human layer?
If only one thing is clear it’s that I don’t know. I have inclinations and intuitions, and a hunch that progressive autonomization is important – especially for service DAOs. To me, this is exciting and what Automatos is all about – exploring.
If I leave the reader with only one provocation, it’s this: what the hell is progressive autonomization?!