WTAF is a DAO?!?!?
May 13th, 2022

A DAO is a Decentralized Autonomous Organization. Well, that was simple wasn’t it? Job done. This has been my TED talk and thank you for coming! Queue the applause.

Except, what does that actually mean? Or more importantly, what is it supposed to mean? What DAOs are today and what DAOs aspire to be couldn’t be more different. At the moment, the majority of DAOs operate in Discord servers with the only differentiator between any other Discord community being that members vote on proposals generated by the organization through a tool like Snapshot where you must hold the organization’s governance token to be able to vote. That’s about as sophisticated as most DAOs are today.

By definition, a DAO should implement a Mode 3 governance model. Great, sounds easy enough, let’s do it! Except, for one problem: Mode 3 governance is very complex. Mode 3 governance is not new either - you may have even been a member of one before in the form of a co-op, or cooperative. According to the ICA (International Cooperative Alliance), co-ops are:

“Cooperatives are people-centered enterprises owned, controlled and run by and for their members to realize their common economic, social, and cultural needs and aspirations.

Cooperatives bring people together in a democratic and equal way. Whether the members are the customers, employees, users or residents, cooperatives are democratically managed by the 'one member, one vote' rule. Members share equal voting rights regardless of the amount of capital they put into the enterprise.”

The first statement sounds very much like how DAOs are often described. The second part? Not so much. “Equal voting rights regardless of capital…”? That’s certainly NOT how DAOs are structured today. Governance tokens derive their value from their voting utility. If you care about the mission of the DAO, you would want to amass more governance token to be able to cast more voting power. Is this right? Wrong? Neither? Let’s back up to the modes of governance before we tackle that thorny question.

Mode 1 Governance

If there’s an “old way” of governance, it’s Mode 1. Mode 1 assigns decision-making authority to individuals. Have you ever worked in an organization where all your work had to be reviewed and approved by one individual? Boom! Mode 1. The larger the organization, the bigger the hierarchy of decision-making individuals becomes. The individual that can make the “final” decision has the authority to accept responsibility for that decision-making process. The quality and speed of decisions, not surprisingly, depends totally on the individual making those decisions. Have you ever worked for a large corporation that had a seemingly infinite number of employee policies? Perhaps the vacation policies were especially complicated. Your manager very strictly enforced those policies. Boo. But your friend in a different department was always on vacation even when they didn’t have PTO! How? Mode 1 governance. The individual managers had the appropriate authority to enforce and interpret the company policies. Different individuals, different interpretations, and different enforcement.

Mode 2 Governance

If Mode 1 is the “old way”, Mode 2 is the “new way”. Where authority in Mode 1 structures is assigned to individuals, in Mode 2 that authority is signed to roles. Instead of a single individual having the decision-making authority, the authority is assigned to a role which can be assigned to a group of individuals. A great example are military ranks. Have you ever seen two military officers of different ranks walk by each other? What happens? The lower ranked officer salutes the higher ranked officer. But they’re not saluting the individual - they’re saluting the rank. Authority increases with increasing ranks. For instance, the commander of a military base might be a colonel. While a particular colonel is assigned to that role, another colonel could also be assigned to be that base commander. The rank of colonel unlocks the ability to serve that role.

While military ranks are hierarchical, agile software development (please stand down with your kanban vs sprints arguments immediately, we are not going there) also implements elements of Mode 2 governance. When software development tasks are generated they are distributed (or taken) by any dev on the team. If you’re a dev on the team, you have the authority to work on any of the tasks. A dev has the authority to implement a task and submit their code. Usually (but not always) a different group of devs or managers would have the authority to deploy the new code to the project. Since the authority is assigned to a group of individuals, when an individual becomes free they can process code deployment in a more efficient manner as opposed to having to wait for a single individual to become available.

Mode 3 Governance

If Mode 1 is the “old way” and Mode 2 is the “new way” then Mode 3 is simply - “the way”. I have spoken. Before we get to what Mode 3 really is, let’s talk about two structures that come very close: matrixed organizations and co-ops.

Are you still there? Did I lose you? For anyone that’s worked at a company with matrix structures, you probably have some hot takes on this topic. Don’t worry, I’m probably about to agree with you on most of what you’re about to say. Large international corporations recognize that decision-making based solely on individuals leads to very slow decisions. In order to improve decision efficiency, individuals often report to multiple teams, departments, or individuals. You don’t have a single manager, but report to different individuals depending on the workflow or project. Remember Peter from Office Space? He had 8 managers. 8 people wanting to know where his TPS reports were. In these matrix organizations, you wear many hats - everyone wears many hats. There are still hierarchical layers of authority, but the number of layers can vary from project, department, team, etc.

Co-ops, on the other hand, don’t have hierarchies. The members have equal say in decisions. Members fulfill different tasks depending on their expertise and experience, but these assignments are decided collectively by all the members. A co-op is a flat structure, with all members having equal input into the decision-making process. While this description of co-ops harkens utopian dreams of equality, reality is often far different. Members contribute depending on their skills and over time, some members will contribute more than others. The members that contribute more might become frustrated that their vote is equal to members that contribute far less - or even not at all. As the co-op grows larger, requiring input from all members can lead to decision-making becoming painfully slow. Some members might suggest that members that contribute the most have authority to make decisions. But at that point, is it a co-op anymore?

Neither matrix organizations nor co-ops represent Mode 3 governance. Mode 3 governance exists at the collision of those two models. Mode 3 is BOTH hierarchical AND flat. Mode 3 uses democratized roles and individual authority. But how?


Matrix organizations and co-ops take almost diametrically opposed approaches to governance structure. Matrix structures create complex org charts with criss-crossing arrows with some people being assigned more roles than they can even remember. Co-ops have a binary org chart - are you a member? Yes or no? The primary weakness of both matrix organizations and co-ops is the amount of human management required to operate the governance model. Matrix organizations quickly become so complex that people can be confused on which tasks to prioritize when being pulled in so many different directions. When you have 8 bosses, whose requests do you fulfill first? A co-op has the opposite problem. Since everyone shares the burden of governance, every member must dedicate time to making tasking decisions on who is the best suited to perform various duties. But what if management isn’t a strong skill - or something that even interests you? As a member of a co-op, everyone must govern for the co-op to function. A matrix organization attempts to compartmentalize workflows so that people with common skills and experience work together while a co-op spreads all workflows across every member. A matrix organization introduces unmanageable organizational complexity while a co-op’s organizational simplicity introduces inefficiency.

A DAO solves both of those problems - well, it SHOULD solve those problems. With a DAO, you get both of these structures combined: members are assigned roles, responsibilities, and authorities like a matrix organization -  but there is no hierarchy and no criss-crossing arrows with a flat structure like a co-op. Everyone has governance input that matches their roles, responsibilities, and authority. But doesn’t this structure introduce the most complexity of all? Yes, yes it does. This structure is totally unmanageable through human governance - but not for smart contract governance.

Without going into deep-dive on smart contracts, smart contracts are software programs that are deployed to a blockchain, stored on that blockchain, and execute on that blockchain. Once these contracts are deployed, they operate autonomously based on the design of the smart contracts and user interactions. Member roles, responsibilities, and authority can be programmed into the smart contracts. Voting rights based on those attributes are coded into those contracts. Decision-making workflows are coded into the contracts. The governance model is a smart contract. Once the governance model is coded into a smart contract and deployed, the workload to manage these governance interactions greatly decreases. There is no question of which boss do I answer to today? Who should vote on a decision? Is a vote even necessary? The answers to these questions are built into the smart contracts.

Even with smart contracts, there are network effects that must be overcome to maintain operational efficiency and scale the DAO. The visual representation of Mode 3 governance is a mesh network of members where the smart contracts can connect members to create workflows based on matching roles, responsibility, and authority. The solution? Clustering. The overall structure is still flat, but members with common workflows can be grouped in clusters where cluster inputs and outputs flow to other clusters. Yes, clustering introduces an additional layer of complexity that once again can be managed autonomously through smart contracts.

There’s only one more problem with what a DAO should be - the name. Decentralized AUTONOMOUS Organizations. Yes, the smart contracts, once deployed, execute autonomously. This attribute is the key enabler to Mode 3 governance. But what are DAOs at their most fundamental level? Organizations of PEOPLE. While the governance model can be coded in smart contracts with decision-making processes managed autonomously through smart contracts, the actual decisions are still made by people. A true autonomous organization would have no people - but would be nothing more than software bots and smart contracts. Governance models must be designed to account for the most error-prone component of the DAO - the PEOPLE. Governance models must be designed with human oversight - checks and balances. Not only are humans imperfect in their decision-making, but humans also write imperfect smart contracts. Will a smart contract ever be perfect? A DAO CAN implement true Mode 3 governance and create a more efficient and democratized organization - BUT a DAO is still an organization of PEOPLE, and must have robust protections against the imperfections of its human members.

InfiniteSn4ke | Mode3DAO

Subscribe to InfiniteSn4ke
Receive new entries directly to your inbox.
This entry has been permanently stored onchain and signed by its creator.