🤷 What’s the deal with DAOs?

Do they work? What are they even for? Should I be afraid of them?

Hi there, welcome to the fourth monthly series of Untangled. If you haven’t yet completed your 2022 New Year’s Resolution, change it to “make Untangled go viral,” smash the subscribe button, forward this email to all the people you know, and then check off that sweet, sweet resolution. Satisfying, right?

There is no Untangled podcast this month. Instead, I want to give you a short selection of my favorite reads on each of the topics I’ve written about to date:

• If you liked “Crypto is not Decentralized” but wanted to dive deeper into the concept of decentralization, check out the first edition of New_Public’s great magazine. This visual guide explores the various configurations of networks, through different degrees of decentralization, and this interview examines indigenous decentralized storytelling traditions.

• If you liked “Some Unsatisfying Solutions to Facebook” but wanted even more cowbell, err middleware, Justin Hendrix of Tech Policy Press and Richard Reisman hosted a thoughtful conversation that included guest of the pod, Daphne Keller.

• If you liked “Let’s imagine an alternative metaverse...” but it left you concerned about the future, I get it, and also, take some solace in the fact that researchers like Timnit Gebru and Lisa Nakamura have launched new research initiatives to address some of these issues. Check them out!

---

🤿 And now for a sociological deep-dive into a topic that you didn’t know you needed to understand: Decentralized Autonomous Organizations.

“Like all great things on the Internet, it started as a meme,” said Jonah Erlich, a software engineer, and one of the core contributors to ConstitutionDAO, which raised nearly $50M over one week in an attempt to buy a copy of the U.S constitution. Yep, that really happened.

DAO stands for Decentralized Autonomous Organization. A DAO purports to be an organization governed by both a transparent set of rules that live in code, and the organization’s members. Technically, a DAO can be formed to do just about anything. The idea is that DAOs are open — anyone can contribute to them. You grant ownership to its contributors by distributing tokens to them; you use blockchains to quickly raise capital across boundaries; you use tools like Discord to enable nonhierarchical coordination across the broader community. I’m italicizing these words because as we know, the language used in crypto often hides more than it reveals. It’s — dad joke incoming — cryptic 😉.

As we’ve seen with ConstitutionDAO, DAOs can be set up for all kinds of purposes. To date, DAOs have prioritized building and funding crypto projects. Remember the 2016 hack of the Ethereum treasury I talked about in my first post and follow-on conversation with Angela? That was the first-ever protocol DAO. But DAOs are increasingly vying to shape the world beyond crypto. KrauseHause, named after Chicago Bulls general manager Jerry Krause, is a DAO that is raising money to buy and manage an NBA basketball franchise. CityDAO bought 40 acres of Wyoming in hopes of founding a city that is governed like a DAO, on the blockchain. There is a DAO working on prison reform, and another trying to address climate change.

Okay, so you might be wondering at this point: why a DAO? What problem are DAOs trying to solve? The self-justifying answer often comes in the form of these two graphs, brought to you by Chris Dixon, a partner at a16z.

What these charts are trying to show is the misalignment that forms between the needs of users and the needs of the corporation. As a corporation grows, they begin to focus more on extracting value from their users, as opposed to simply attracting users — this explains Meta’s (and others’) constant need for your attention and data.

People like Dixon insist that DAOs solve this misalignment problem: the corporation is taken out of the equation entirely, and economic/governance rights are put in the hands of the organization’s members in the form of tokens, which is supposed to align the interests of the individual contributors and the organization. The rules that govern the members are encoded into a blockchain, which means the DAO cannot stray from its original purpose.

Except, DAOs don’t really solve the misalignment problem that they claim to solve.

Rather, you just get new forms of misalignment. Here’s the problem: publicly-traded tokens turn the communities into markets. Tokens financialize everything they touch. This isn’t necessarily a problem if the mission of the DAO is to generate returns, like in the case of FlamingoDAO, which invests in NFTs. But what about those with social goals? Let’s start with some very apt observations from Danny Zuckerman, co-founder of 3Box:

“Communities are full of deep intrinsic motivation and hidden order. Not all of them can be plotted, attributed, measured. Tokens can add legibility and direct incentives. But they can also wipe out far more powerful organizing principles in the process. [...] The nightmare would be that DAOs do rise, only to accelerate the dominance of shareholder return over all other values. They are efficient and open to all — but for what? What makes sense for financial networks is not necessarily good for non-economic relationships.”

Taken from this Twitter thread.

This poses a problem for DAOs that claim to exist in order to build a community: if your ‘community’ introduces a publicly-traded token into the mix, some members (new and existing) will find that their motivations shift from ‘building a community’ to simply trying to make money — and there’s the misalignment. This table I drew up demonstrates where misalignment occurs in different types of DAO:

DAOs only really avoid misalignment when the sole purpose is financial gain, OR when the DAO is community-driven, and no token is introduced.

ConstitutionDAO is a perfect example of a DAO with a mixed mission, and therefore one that has suffered from misalignment. Some members wanted to co-own a piece of history (err, or rather... have the ability to vote on where it was displayed), while others simply wanted to get rich by eventually selling the token. Friends With Benefits is a DAO that functions as a kind of social club for musicians and artists, and already, it has become “a victim of its own success,” as the kinds of members it wants to attract can’t afford the token-based membership fee.

This kind of ‘token as ownership’ system means that those with the most resources will likely be the ones making the decisions in the end (unless, of course, those with the most tokens encode governance principles that reduce their power over time 🤞.) Whether it’s how to compensate people, resolve conflict, or which projects to fund — those with the most tokens will often have the most say.

DAOs face the same problems as other flat, decentralized, organizations.

Many DAOs claim to be decentralized, flat structures; they use blockchain technology and other tools to coordinate in a nonhierarchical fashion. The problem for DAOs, as posed by decades of research, is that the removal of formal structures does not actually level the proverbial playing field — it only leads to informal structure and hierarchies. Authority and control aren’t represented by your boss telling you to do things. Rather, informal relationships, reputation, culture, and moral codes take the reins.

💡We also know that people forge networks and social ties based on social similarities. A process that lets anyone join may actually lead to homogeneity. Anyone who works on DEI in an organization will tell you that you have to proactively recruit to identify a diverse applicant pool. It’s not about letting anyone who wants to join do so (re: equality) but treating some groups differently to achieve a more equitable outcome (re: equity)

With ConstitutionDAO, the memes themselves became the mechanism for cohering a group identity and shaping group behavior. As Packy McCormick wrote in his newsletter Not Boring, the group used Nic Cage memes, distinct iconography, and the rallying cry “we’re all gonna buy the Constitution” (wagbtc) to spread the message and raise funds. He adds,

“These memes, acronyms, and symbols create cohesiveness within the group, a bridge to other communities, and a way for members to identify themselves to each other across the internet and reinforce each others’ enthusiasm.”

It turns out, replacing formal structures with informal ones leads unsurprisingly to exclusion and inequity. In “The Paradox of Meritocracy,” Castilla and Bernard write that “the meritocratic beliefs lead individuals to feel unbiased, fair, or objective, and as a result become more likely to express individual bias toward low-status groups.” As a result, research has found that across the board, women and people of color tend to do better in bureaucratic organizations than “flat” ones. This is also why the use of tools like Coordinape and SourceCred, which help DAOs measure and reward individual contributions, deserve scrutiny — they are likely to reproduce gender and race-based inequities under the guise of meritocracy. In the limited empirical research on DAOs to date, contributors are already expressing concern over the “lack of social equity and diversity in most DAOs,” and the problem of “misogynist people.”

But if a given DAO is indeed ‘flat’, it probably shouldn’t be ‘autonomous’.

DAOs are said to be autonomous, in that they operate through the use of smart contracts and programmed technologies, which creates less need to make deliberative decisions — because many of those decisions are already pre-determined in the tech itself.

But the literature on organizational sociology says, that with flat structures, legitimacy comes from the decision-making process. So if you’re going to be flat, your decision-making processes need to be water-tight and seen as valid. In The Conversational Firm, professor Catherine Turco talks of voice rights, which refer to who gets to speak and have their voices heard. She argues voice rights are often seen as more important than decision rights, which refer to who gets to decide what:

“Studies of procedural justice in the legal system and inside firms have found that the process by which decisions are made is often as important to feelings of fairness and justice as are the decisions themselves.”

Catherine Turco, The Conversational Firm

Essentially, what Turco is saying is, if the process of decision-making is itself seen as legitimate, there can be disagreement on the ultimate decision without it leading to conflict and mistrust.

Unfortunately, many DAOs haven’t put a lot of thought into how they make decisions. They often operate by “rough consensus” which helps them “move fast”. ConstitutionDAO did exactly this, and their decision-making process was, to put it mildly, not seen as legitimate. In one Discord chat, a frustrated participant to ConstitutionDAO wrote:

“IMO, this project lost the community the moment publicly made announcements were deleted. It created speculation and uncertainty could have been avoided. It’s leaderless chaos. Worst of both DAO and real world. people had $people token, why weren’t they asked to vote?”

A revealing screenshot from the ConstitutionDAO Discord server.

So if rough consensus doesn’t work, how should DAOs govern themselves? To avoid the horror of the above, some DAOs are already shifting towards a delegated authority model. In doing this, they’re sort of admitting that making decisions in a networked community is incredibly difficult. A shift towards already existing governance models also throws into question what makes a DAO different from any other flat/decentralized organization. It’s clear that, at the moment, DAOs struggle with building decision-making processes that their members consider to be legitimate, so conflict and mistrust do often bubble to the surface.

Of course, those championing DAOs argue that they replace the need for trust with “radical transparency”. Mario Gabriele, who writes The Generalist, described the relationship between transparency and trust this way:

“Transparency plays a crucial role in any organizational culture as it establishes trust between all players. DAOs enable this at an unprecedented level due to the public and immutable activity of any DAO’s ETH address on the blockchain. This creates both an implicit and explicit checks and balances mechanism that allows for a community of stakeholders to both stay informed on how a DAO is exercising its capital while ensuring a leadership team is making decisions in line with the community.”

But, as scholars like Primavera De Filippi have argued, blockchains are not “trustless.” Indeed, trust is social and relational; it doesn’t result from making all information transparently available. Indeed, trust that requires complete information sharing isn’t ‘trust’ at all. Trust researcher Charles Feldman defines trust as “choosing to risk making something you value vulnerable to another person’s actions.” This isn’t a one-off proposition, rather, it’s built over time. Trust looks like a boss regularly owning their mistakes or believing that a friend will pay you back (because they always have done in the past).

DAOs are still very much an experiment...

ConstitutionDAO was ultimately outbid by hedge fund CEO Ken Griffin. Contributors paid significant fees as they scrambled to get their money back. It was a mess, basically. But, it captured the public’s attention in part because it feels like representative democracy is slipping through our fingers; there is a general feeling that our social and political systems are failing us, so it’s no surprise that there’s an appetite to pick up the slack — this is why we’re seeing DAOs that focus on climate change, or under-investment in black female and nonbinary entrepreneurs.

Unfortunately, right now DAOs are falling into the trap Meyer and Rowan address in their research, namely that “formal structures of many organizations reflect the myths of their institutional environments instead of the demands of their work activities.” In other words, the myth of decentralization and autonomy are themselves encouraging a decision-making structure that looks like “leaderless chaos.”

DAOs are an experiment in how we organize ourselves. That’s fine — we should try new models. But they aren’t a radical innovation in organizational structure. Supposedly decentralized and/or flat organizations have been around for a while now. And the introduction of publicly-traded tokens will impact the organization’s mission and alter the motivation of those contributing, raising the question: will DAOs innovate in governance if it means that those with power and money have to give it up? That would be truly innovative.

Thank you to Nathan Schneider for providing comments on an earlier draft and to Georgia Iacovou, my editor.