This is a cleaned-up version of a Discord thread I posted between 2021/11/06 and 2021/11/09. The intent with posting this here is to make this material easier to find, reference and share as appropriate.
Wanted to inject some ideas into the conversation in a not horribly distracting way, so I am hoping that using a thread will accomplish that.
First a disclaimer: I’m still learning a lot about DAOs and ignorant of a lot of work already done in this space. Don’t hesitate to point out where I’ve ignored obvious knowledge sources. Suggestions, arguments, and questions are welcome and all input will be considered - some longer than others.
Let’s start with getting clearer about what we mean by “governance of DAOs”. When I talk about governance of DAOs, I’m focused on value and how we maximize it. While this is typically meant in financial terms, I mean value in the broad sense of something people care about. Money is one way to measure value, but it’s not the sole means.
Most references to governance miss this, and it causes a lot of issues. In traditional organizations, governance is equated with committees, bodies, etc. In much (but not all) of what I have learned about DAOs so far, governance is equated with voting. These ideas aren’t wrong per se, but they are limited in how much they actually help increase value.
The best formulation of value I have seen comes from a group called ISACA and their COBIT framework. They break value down into three components - benefits, resource costs, and risks (BCR). Ideally, we get more benefits than we expend in resource costs, with a level of risk we can stand. Again, this is about more than finance. The benefits may be in well-being, the resource I am spending may be time with my family, and risks may be to a relationship. Whatever matters to us.
COBIT also calls out 3 aspects of governance - evaluate, direct and monitor (EDM). While not as fun as electronic dance music, it is a useful cycle. We evaluate our current state of value, direct actions to increase value, and monitor changes in value that lead to further evaluation. All 3 matter.
This highlights that the core of good governance is effective decision making. This is where concepts like John Boyd’s OODA (observe, orient, decide, act) loop are helpful.
What makes for a good decision? Well, that’s a value question, so we can use BCR to answer that. An ideal decision provides the best benefit/resource tradeoff, without delay or coordination costs, with no risk of failing. In practice that’s impossible, of course, but the closer the better.
When I look at DAOs through this lens, it’s obvious that we often fall far short of an ideal decision. Members may not agree on what benefits they want, what resources they are willing to expend to get those benefits, and what their risk tolerances are. When everyone has to vote on everything, that has a cost and creates delays. These factors alone increase the risk of failure, and that is magnified when you consider that it’s hard to get 100% commitment to a decision that only 51% of people wanted.
This also starts to shed light on why scaling is so hard. The evidence is in - scaling doesn’t work. Anyone in a traditional organization that has grown knows that an 8 person organization can’t function the same way an 800 person organization does. An insect scaled up to human size would collapse under its exoskeleton. The 50 foot woman literally couldn’t be human at that size - her bone density would not support her body. Geoffrey West’s book Scale includes many great examples of these challenges.
We often use the word “scale” for “growth”, but they aren’t the same. Nature doesn’t grow by scaling, it grows by reproduction. The promise of DAOs is that they can harness the principles of reproduction far more effectively than traditional organizations.
This is the main reason that I was attracted to Orca Protocol. Pods have the potential to enable DAOs to coordinate on the big picture effectively while minimizing the costs of doing so. But to do this, we need an approach to organizing that can take advantage of the power of pods.
We have to actually design support for growth into DAOs at their start and make governance part of how they work from day 1. But again that doesn’t mean subcommittees or just having everyone vote on everything. It does mean making value clear in BCR terms and promoting good governance hygiene with minimal burden.
To make value clear, a notion from Joseph Campbell’s Hero’s Journey is useful - the call to adventure. The best way to get to a common view of value in a DAO is to have a shared vision that guides everything the DAO does. But if we want this to scale, by definition that vision has to be bigger than what the DAO can ever fulfill. Otherwise, we top out and lose our way.
Vision statements get a bad rap in many circles, and deservedly so. However, the concept is still helpful if we approach it the right way. The best approach I have seen is by a woefully underappreciated thinker named Tom Graves, who talks about the concept of “Vision, Roles, Mission, and Goals”. Building-blocks for a viable business-architecture – Tom Graves / Tetradian
To reduce the risk of lousy vision statements, let’s coin the term “adventure-vision” to apply Graves’ concept of enterprise-vision to DAOs. As he puts it, the adventure-vision describes a desired world, ideally no more than half a dozen words, about the ‘what’, ‘how’ and ‘why’. I love how he describes it as a way of saying "this is what interests us – and if this interests you too, perhaps you should speak with us“ A structure for enterprise-vision
A key point is that this adventure-vision DOES NOT CHANGE. It serves as an attractor that enables a common orientation as we grow into multiple pods and as the world changes around us. Now the link to decision making is clear via OODA. A common orientation helps us better interpret observations, make decisions, and take action.
Creating a DAO without being clear on the adventure-vision will lead to failure. That’s one of the reason why financially-oriented DAOs may tend to hold together better. In that case, the adventure-vision is more consistent since getting rich is pretty interesting!
The DAO then needs to identify the role(s) it will play in the adventure-vision. What will it do, and not do, to bring about that world? This is where strategy is useful. Again, the term is horribly abused, but I like Richard P. Rumelt’s notion that he shared in “Good Strategy, Bad Strategy” that can be roughly summed up as a diagnosis, guiding policy and coherent actions.
Diagnosis can be thought of as what is preventing the adventure-vision from being real. It’s a simple, understandable story that suggests the role(s) the DAO might play in bringing the adventure-vision to life.
Guiding policy in a DAO would primarily focus on setting constraints from a risk and resource cost perspective. Just like in nature, we must survive long enough to make an impact or pass along our genes. This provides the “not do” that complements the “do” from the diagnosis.
Coherent actions come from the diagnosis and guiding policy. The more the DAO’s actions support each other, the bigger the DAO’s impact will be on the adventure-vision.
These coherent actions take the form of missions. Yes, mission statements are terrible, but the concept as Graves describes it is helpful. A DAO’s mission(s) are the capabilities and services it intends to create and maintain to fulfill its role(s) in service to the adventure-vision.
This creation is guided by goals. A goal in Graves’ approach is a project - a set of deliverables with a target date for completion. As goals are accomplished, they create services and capabilities that the DAO then operates as part of its mission. Ultimately, accomplishing these goals is how we bring the adventure-vision to life!
So just for fun, let’s try to describe what Orca Protocol is doing using this approach and see how badly I misunderstand things.
The second paragraph of the “Closing Thoughts” section of “The Eightfold Path to DAOism” post fits very well with this structure! The Eightfold Path to DAOism — Mirror
Orca Protocol’s adventure-vision: “Help DAOs achieve their full potential”
Its role: “make governance accessible by creating tools around a DAO’s most basic primitive: people.”
Its mission: provide “a modular and flexible body to manage participation, shared assets, and organizational permissioning”
Its goal: create a pod primitive that “allows for dynamic and composable structures to be created around any party of actors within a DAO ecosystem, while introducing mechanisms for accountability, incentive alignment, and checks and balances”
Obviously the goals will change a lot, the missions will expand over time (hopefully!), and even its role will evolve. But the adventure-vision will remain.
So the approach seems to check out, which provides at least a bit of credibility. But how many DAOs can actually state their adventure-vision, roles, missions and goals? And how closely would their members match up if you asked each of them separately? I suspect there is a strong correlation between the gap between members and the health of a DAO.
Let’s talk about decision making some more. Specifically, decision capacity. As in, it’s limited.
The corporate world tends to forget this, which is why in large orgs senior leaders spend all of their time in meetings (governance and otherwise). Unless they are deliberate in carving out time for strategy and planning, it’s very easy to fall into reactive mode. It’s frankly remarkable that this works as well as it does, but it leaves a lot of potential on the table.
In a DAO, this can be even worse. If everyone has to vote on everything, I have to choose between the following options: spend a lot of time on research, vote ignorantly, or miss a lot of votes. Of course, if we naively scale, we can do all 3!
So what’s the right amount of decision capacity to invest in a decision? Of course “it depends”, but there are a few things we can say with certainty. The first is that if there is a best answer, any decision capacity we invest past the point of feeling confident it is the best answer is wasted. So we want to be efficient here too. If we have too many decisions to make, we want to invest our decision capacity where it provides the most benefits versus the resource costs.
The next is that not all decision capacity is created equal. I can waste a lot of time trying to figure out which medication you should take for your ailment, but a doctor is much more likely to provide both the right answer and do it more quickly. Of course, if a specific decision maker’s capacity is committed to higher value topics, we may have to settle for someone less efficient, or accept a higher risk of a bad choice.
The decision approach also has a big impact here. This is where the remarkable Forrest Landry’s piece “On the Nature of Human Assembly” is so helpful. https://www.magic-flight.com/pub/uvsm_1/sgrp_small_group_2.pdf
Landry posits that all decision making processes can be described by three types, which I’ll quote in full since this is so important to understand:
“In a democracy, a range of possible options is reviewed, discussion and persuasion (rhetoric) is followed by a vote, and the majority decision applies to the whole.”
“In a meritocracy, some process (it does not matter which), is used to select a focus of decision making (a single person or smaller group), which then (perhaps after listening to the considered council of others) will make a decision which is applied to the whole. It is assumed that the whole will always and implicitly trust the part to make decisions ‘correctly’ (whatever that notion is taken to mean for the whole group).”
“In consensus, all members sit in council together and discuss potential decisions together until a complete and total agreement is reached among all members (however long it takes) which then becomes the decision of the group.”
Democracy, meritocracy, and consensus. They can be mixed and matched, but these are the composable building blocks of decision process.
Each has its strengths and weaknesses. As Landry points out, democracy is faster than consensus, and less prone to abuse than meritocracy. Meritocracy puts more power behind a decision than democracy, and is more adaptive than consensus. Consensus strengthens bonds between members more than democracy, and is more likely to avoid blind spots in decision making than a meritocracy. Yes, there are edge cases, but in general this isn’t telling you anything you don’t already recognize as generally true.
The real genius of Landry’s work is to recognize the differences in each approach and combine them into an overall approach that gets the benefits of each approach while minimizing their weaknesses. I recommend reading his whole work, but my takeaway of how this would apply to a DAO is as follows.
Defining the adventure-vision, role(s), and mission(s) of the DAO should be performed by consensus. These define what the DAO is, and there are always good returns from a better answer to these questions. The power of a true believer is hard to ignore, whether it’s of a religion, crypto, or crackpot conspiracy theories (are those 3 things, or just 1? )
Executing against goals should be performed by meritocracy. When Vitalik codes and I make dinner, we get Ethereum and a nice meal. If I code and Vitalik makes dinner, we get Hello World (on a good day) and the food might be inedible too (he’s never cooked for me, so maybe he would best me there too?) This is the whole comparative advantage notion of David Ricardo and is pretty well established.
Curbing waste and abuses should be performed by democracy. That probably sounds wrong to you if you live in the US, but hear me out. People can much more easily identify what they don’t want than what they want, as anyone who has ever met their friends for a meal can attest to. Thus, if we’re getting nowhere in trying to decide on whether to adopt a new mission, democracy is a great way to get us out of that rut. Once enough people get frustrated and vote to stop talking about it, we can move on to something else. Likewise, if the person in charge of the project turns out to be an incompetent klepto monster, we vote zim out and pick someone else.
Exactly how to put this together into a DAO playbook is something I am playing around with, yet have not yet fully figured out. But the idea is to put something together kinda like Holacracy, except that people can actually follow it.
The main reason I bring this up is that Orca Protocol needs to have the ability to support all three decision methods (and their checks and balances) in order to make this work. If a new DAO starts its life with a primitive that enables this balanced method to work, it will deliver higher value against its adventure-vision than alternative DAO approaches (and conventional organizations, to boot).
Of course, success will likely mean growth, so how do we deal with that once the DAO gets anywhere close to Dunbar’s number? More to come on that.