DAOs: From fractal primitives to network scale.

This article was originally published on superbenefit.mirror.xyz

Rowan is a contributor to SuperBenefitDAO and works on scalable models of coordination and fractal DAO governance. Deep gratitude to Heenal Rajani and Michael Lewkowitz for contributing to this article. This article is the third in our series exploring the concept of DAOs as networks and their evolutionary nature.

Scale matters

DAOs are extraordinary for their capacity to test experimental ideas. And the innovation DAOs are unlocking because of this, all across Web3, is astounding! But, in order to have a transformational impact on the world, we need DAOs to scale beyond experiments. There is a risk that if they can’t successfully scale what they do, their potentially transformational innovations will just get co-opted by more centralised organisations (pseudo-DAOs and companies), who can more effectively scale and exploit them. Leaving us in much the same position we are in now.

Scale is not just about getting a lot of users onto a protocol with a bare bones interface that only early adopters will accept. In traditional startup speak, this would be considered validation or proof of concept only. Scale means delivering world class products to millions of people, with user experiences that are better than what Web2 is currently giving us.

I know we like to criticise the Web2 progenitors to our world, but these companies are running massively scaled operations the likes of which DAOs have not come close to yet.

It is true that digital scaling (network effects + scaling at near zero marginal cost + Moore's Law) means that digital business models scale in particularly powerful ways. But this does not mean that these models scale effortlessly without needing infrastructure. The likes of Google, Meta, Twitter et al still have to design and build significant infrastructure to generate scale advantages for their business models across; design, software development, sales, marketing etc – just like any other business does. And likewise, DAOs don’t get to opt out of this dynamic just because they are digitally native.

It is also important to remember that DAOs are increasingly going to be building for real world applications. If Web3 is going to be a transformational force in the world, DAOs are going to need to operate across all of the domains that companies currently do. And for this to work, being able to harness scaling dynamics at least as well as companies currently do is going to be crucial.

Scale needs to be a clear part of DAO thinking going forward and we need to make sure that we aren’t ignoring it – consciously or unconsciously – just because it reminds us of traditional business.

In the previous articles in this series – “DAOs aren’t things… they are flows” and “Scale and the levers that provide DAOs their power” – we looked at how a legacy mindset causes us to think of DAOs as ‘entities’, but they are much better understood as networks with coordinated flows of resources and value. We then unpacked how biological systems, companies and cities all function based on two fundamental scaling dynamics:

We saw that unlike companies that operate business models, DAOs are network economies. This makes them more like cities, giving them the remarkable ability to leverage both Dynamic One and Dynamic Two scaling systems.

In this article we will dig deeper into how DAOs can deliberately design themselves to utilise both of these dynamics.

From simple to massively complex

DAO structures need to start simply but then be able to grow into massive complexity. As DAOs try to scale, we cannot just apply arbitrary structure to them to solve the coordination issues that will naturally show up. If we do, we are just going to end up with broken systems. Ad hoc solutions that seem like good ideas at the time will end up causing problems like centralisation and fragility down the line in unexpected ways as they scale.

Galls Law reminds us that we need to start with simple systems that work:

SuperBenefitDAO recently hosted a conversation on metagovernance with Gabagool.eth, BPetes.eth and Justinehy.eth. As we explored the topic of metagovernance, one theme that came through strongly was the idea that we need to keep coming back to the magic that we all saw in DAOs in the first place. The magic of small groups of inspired people working together on something that matters to them – and the simple primitives that make this possible. From here we work our way up to complex governance and metagovernance, but we keep this simple starting point in focus. This is a great signpost for us to think about the simple primitives out of which complex DAO networks can emerge.

DAOs as fractal networks of connected cells

When a core team of people come together to start a DAO, pretty quickly they feel the need to create sub-teams as a way of separating tasks and managing accountability. As soon as this happens, they have begun the complex and difficult journey of experimental org design.

The mental model that rescues us from this becoming a morass of incoherent and unscalable structure is that of fractal design. Rather than umbrellaing these teams inside an entity (like a company does), we can think of these smaller teams as being identical in type to the DAO that birthed them. Each starts as a DAO primitive; a version of the DAO created out of itself, in order to get things done. All of these DAO cells exist in relationship with one another in a complex and emerging network.

Call them subDAOs, Pods, Fractals, Working Groups, Circles – I’ll refer to them as “cells” in this article because that works nicely with a biological metaphor, but the naming of them is neither here nor there – the mental model is to understand them as their own independent DAO cells that are defined by their relationships with other cells in the DAO network.

This notion of DAO cells as the primitive starting point for DAO networks is the basis for DAOs being able to leverage both of the scaling dynamics laid out in the previous article.

It is also the mechanism by which DAO networks collectively evolve.

Here is how both of these mechanisms work.

Cells can become Scaling Hierarchies AND Creative Networks

When a new DAO cell is spun up in a DAO network, depending on the purpose it has, it needs to be set up to be either part of a scaling hierarchy or a creative network.

D1 Scaling Hierarchy: If it is to play a role in a scaling hierarchy – for example, as part of DAO operations or core DAO infrastructure – then it will have quite tightly defined accountabilities. It will be closely coupled to other cells in the scaling hierarchy because its outputs will be critical to the functioning of the system as a whole.

D2 Creative Network: However, if the cell came into existence to explore things like novel ideas, partnership opportunities or new areas of research, then it will be much less constrained.

Where a cell needing to plug into a scaling hierarchy needs tight constraints to make sure that it is producing the outputs that the other cells in the hierarchy need, a creative network cell is set up to be maximally unconstrained.

DAO cell primitives need to be easily configurable in order to dial up the right constraints to make them function as the DAO network needs. This requires clearly defining the accountabilities, roles and relationships with other cells in the network, and then setting up the corresponding governance structure and tooling to make this possible.

Out of these single-celled starting points, DAOs then evolve into complex networks of cells.

Hats off to all the folks already experimenting in sophisticated ways with how these primitives can work – Orca, Sobol, Colony, Murmur, Juicebox, Hypha et al.

Photo: Matthew Ansley - Unsplash
Photo: Matthew Ansley - Unsplash

Photo: Matthew Ansley - Unsplash

Example: DAO Prison Network

Let's make this more concrete by looking at a hypothetical example. Imagine a national prison system run as a DAO network. I’m choosing a difficult real-world problem here as an example to emphasise the challenges of scaling real-world systems.

The prison network is animated by its purpose of transforming how we approach crime as a social issue by searching for better solutions to these complex problems.

Because it is running existing infrastructure in an existing system, the DAO network would be made up in large part by scaling hierarchies consisting of networks of DAO cells that combine to deliver the well understood operational needs of the prison system. Things like buildings and facilities, security, health and welfare, education, staff/HR etc. These hierarchies of cells would be highly coordinated to deliver the right outputs, reliably and efficiently.

Each DAO cell would be set up to perform a specific role in the system, be accountable to other cells that rely on it, and be governed to some extent by these interdependent cells. They would likely have ongoing streams of funding and would measure themselves by the effectiveness of their activities in delivering on the known outcomes they were created to produce. They would also have the ability to produce their own new DAO cells so that as their responsibilities grew they would be able to scale their activities via fractal hierarchical scaling.

An important caveat to add in case the word hierarchy is triggering. The term scaling hierarchy refers to the structure of a network that functions with this particular set of scale advantages. It doesn’t need to mean “dominator hierarchy” or “management hierarchy”. Companies traditionally use management hierarchies to implement a power structure, but DAOs don’t have to do this. Much of the cutting edge work in organisational design (The Ready being a great example), concerns itself with how we can coordinate at scale without resorting to old school management hierarchies, aka “command and control”. But make no mistake, DAOs do need to match their ability to innovate with a genuine ability to scale.

But interspersed with these scaling hierarchies would be numerous loosely coupled creative network cells that are experimenting across the system in search of better solutions. Unlike the more highly coordinated scaling hierarchies which function with a well defined operating model and something that looks a lot like “strategy”, these creative cells will bubble up from ideas and inspiration from the DAO’s community.

A member of the community can have a novel idea, refine it, get a proposal approved and then spin up a cell to start exploring it. For our prison example, things these cells might experiment on could look like:

  1. Ideas for better systems for staff - Process innovation/Horizon-1 (in traditional business speak)
  2. Novel ideas for working with prisoners, rehabilitation etc - Product line expansion/Horizon-2
  3. Radical ideas that would make prisons as we know them obsolete - Disruptive innovation/Horizon-3
  4. Ideas that bridge and extend the DAO network beyond prisons, perhaps into mental health or early social crime prevention. Possibilities to connect and integrate the prison DAO into larger aligned networks beyond prisons.

These creative cells would be funded and structured to give them maximum freedom/autonomy. And then as creative cells evolve their ideas beyond experimentation, they can grow and connect with other parts of the larger DAO network to develop their discoveries into scaling hierarchies that the whole network can benefit from.

DAO networks evolve organically

With traditional organisations and companies we are used to thinking about power structures. And so for an organisation to change, it requires the power structure to change at the highest level and then for this to cascade down through the hierarchy. However, because DAOs consist of networks of cells, they can evolve organically.

For example, if a creative cell in our national prison DAO had been exploring an idea for early social intervention which prevented crime and discovered an exciting solution that would meaningfully reduce crime. That cell could generate its own contributors, community, token value, capital etc and as it grew would exert its own gravity on the larger DAO network. As this new opportunity started to scale it would change the shape of the original DAO as the network collectively reorientated itself around this new and impactful opportunity. Perhaps as a result the prison population would fall, but this is good news as the DAO would already have been flowing resources towards the new activities and possible new paradigm of transformational social outcomes.

This is sense-making in action. Rather than leaders in a power structure gathering up all the data and deciding on a new strategy for the DAO, the energy and resources flow with each individual member of the network as they gravitate towards the highest potential and most aligned opportunities. This is the DAO network reshaping itself organically using genuinely decentralised intelligence. Rather than trying to define and implement strategy for the DAO, strategy emerges autonomously from the DAO network itself.

This is in stark contrast to traditional organisations who can’t produce this kind of flexibility and collective intelligence due to rigid scaling infrastructure (see previous article). This legacy model leads to incredibly crappy externalities. This is easy to see in current private prison systems, where prison operators will often lobby governments for stricter enforcement of laws to increase prisoner numbers, in an attempt to maintain occupancy rates and ensure their business models keep working.

Enough scale to pay for the innovation

These amazing network dynamics are great in theory but DAO networks need to generate enough scale to make them work in practice. Searching for novel solutions is resource intensive. If DAOs are going to be these combinations of scale and innovation that can sense and evolve, they need to be big enough to generate the scale efficiencies to afford to do it.

Scale is also needed to make sure that these valuable new ideas make it out into the world in a way that many people can benefit from them.

It is easy to see from our prisons example, that if the DAO network only included one prison, the system would likely not be able to sustain the kind of creative networks needed to actually find breakthrough solutions. And then if they did find a breakthrough solution, they would not have a network through which to scale it. In our previous article we talked about the city-like positive feedback mechanisms that DAOs can generate:

For all of this to work the DAO network needs to be operating at the scale of the system that it is trying to transform.

DAOs as many interconnected networks working across vast scale

But all this talk of “enough scale”, must not collapse back into entity thinking. It is not about any one particular DAO being any particular size. This is the exciting opportunity. If we can pull the design of DAOs back to their smallest starting primitive cells, and DAOs (as networks of cells) just build up from there, then getting bogged down in where one DAO finishes and another starts becomes unnecessary.

If one DAO cell can easily, and in a principled way, connect with any other DAO cell and form into a useful structure (scaling or creative), then in principle it should be just as easy for two cells in different DAO networks to connect as well. Given all the experimentation with things like token swaps and DAO-to-DAO tools, the technical and governance tooling to do this more and more easily is emerging.

In this way I don’t think the future will consist of giant monolithic DAOs that rival the likes of Google, Meta and other giant traditional companies. DAO networks present us with a different way of thinking about scale. A company needs to contain its hierarchical scaling dynamics inside an entity, but the tools that Web3 gives us allow the coordination of resources across and between vast networks.

So a Web3 solution that replaces Google might be a whole web of DAO networks that is spread across all of Google’s 4.3 billion users… some of whom are working in highly structured ways to deliver at-scale technical infrastructure and core functions, some who are working on thousands of smaller experimental projects that grow and evolve into new and valuable things, some who just use the products being created and contribute data to the network as a byproduct. Maybe there is one search product, maybe a hundred, maybe a million.

I know developing a million products sounds implausible but there is no reason that the tools we are collectively evolving to coordinate shared ownership, governance and activity across interconnecting networks could not scale to coordinate a network of billions.

DAOs are “network intelligence” that flows

DAO networks will find equilibrium as they scale and fill the needs/opportunities they are trying to fill. But they will constantly be evolving under the surface. Whether it is the culture of the network evolving within the larger cultures we inhabit; adapting to changing environments, developing new technical underpinnings or product/service offerings, harnessing emerging discoveries and innovation - they will be able to scale but also to evolve, grow and reorient towards what is needed.

Unlike animals and companies who, because of their inflexibility, grow old and die, DAO networks will be like forests, dying and being reborn moment by moment as the energy and resources of the community flow, form, reform, connect and interconnect in ever more complex and useful ways.

But DAOs developing into these kinds of complex emergent intelligent systems can’t happen in an ad hoc way, it has to be foundational. We need to focus on finding and refining the primitive starting points. We need to make sure that we are working towards simple systems that can evolve into vastly complex ones.

The limited liability company primitive was able to produce the vastly complex global systems we have today. Now we need DAO primitives that can similarly evolve into our next emerging paradigm.

Subscribe to Rowan Yeoman
Receive the latest updates directly to your inbox.
This entry has been permanently stored onchain and signed by its creator.