/ Research

Aragon Labs Research Update #1

Latest news from the Aragon Research & Development department, Labs

Our research efforts are on exploring how to realize the lofty ambition of becoming a decentralized, digital jurisdiction. Both the scope and range of services a jurisdiction could offer to attract its constituency are vast and worth exploring. The single most crucial consideration right now is how it will be governed. With effective governance processes we can create new tools for the community. Tools to create new services, or replace or alter existing services, as required to make organizations more effective.

Governance processes enable us to address common organizational challenges such as the following: How do we deploy infrastructure upgrades over time? How do we ensure these upgrades are sound, both on the technical and social aspects? How to divide resources to maintain and deploy new infrastructure? How do we manage resource allocations to ensure stability and self-sustainability long-term?

These questions are generally quite simple to address in the context of a traditional organization. But are much more complex when it comes to Decentralized Autonomous Organizations. In a traditional organization, responsibilities are defined in a subjective language. The organization inherits powerful concepts like fiduciary responsibilities from a long established legal system. A system with a body of law and the capability to enforce those laws with force if necessary. In contrast, the rules of a decentralized autonomous organization are all code. Codified in an immutable smart-contract at the time of instantiation. And they cannot be altered except by acting within the organization's own self-defined rules. Its constituents are as likely to be other pieces of code as they are to be people that have bank accounts. Entities that can get thrown in jail. Thus, it is critical that we research and experiment with care. And use different codified governance processes. These we need to do to find solutions that align the interests of all participants.

There's no substitute for experimentation. It's a requirement for being confident in a process' resilience in the face of practical application. As part of our Labs initiative we want to experiment rapidly and in parallel with different processes. But we also want to ensure that we do so in a resource efficient manner. This means starting with small scale experiments, seeing what processes are most effective. Then apply them to more significant applications after our confidence has increased in the process due to the Lindy Effect.

Token Curated Registries as Governance Process Experiments

The canonical version of a TCR is an example of a specific governance process intended to create intrinsic incentives around the curation of a list of arbitrary items. It has garnered tremendous hype and that hype has outpaced the implementations. It has been referred to as a cryptoeconomic primitive or a building block, despite being largely untested in practice. And even when it has been tested, it has shown that the model still needs improvements to be practical. This is not to say that TCRs don't work or won't work (I think they will). But we need to take a step back and recognize that we need to experiment and test our assumptions thoroughly. There is a silver lining, bear markets seem to be good for that sort of thing!

This emphasis on using TCRs as a vehicle for experimentation in governance processes is what motivated us to work on a Modular TCR implementation based on the Aragon tech stack. Our implementation consists of four distinct Aragon Apps. While still a work in progress, if you are technically minded, the contracts are in a functional state. And if you're familiar with how Aragon DAO kits work, you may enjoy playing around with the current version of the basic TCR DAO kit. Keep in mind it is still under active development and things may be buggy, lack coherent documentation, or even a front-end at this point. Of course, if you're interested in contributing, don't hesitate to do so!

As our TCR implementation matures it can be leveraged in several important ways.

As an proving ground for governance processes

Registries and Lists are useful in many organizational contexts. Within the Aragon ecosystem there's Aragon Package Manager registries and our community-curated newspaper Aragon Monthly. Even a curated list of DAOs would provide utility for the community. These use cases may not be feasible as independent TCRs because they do not necessarily represent a large enough independent economy. In this case we can choose to subsidize the curation process to evaluate different governance processes while providing utility to the ecosystem.

As a critical part of the Aragon Network

Core aspects of the Aragon Network may need a curated list (or more generally a curated data structure). For example to enable ANT holders to govern the Aragon Court's hierarchy. Another possibility is to use a curated list to provide a subset of users special roles and responsibilities within the network. Such as a technocratic council that is granted veto privileges for proposed infrastructure upgrades.

As an early DAO compatible business model

Early on the business models that make sense to operate using a DAO are those that are blockchain native. There are several projects which are using a curated list as a core element of their business models. Whether the organizations wants to evaluate the dankness of memes. Or to organize a mesh network. The ability for organizations to take advantage of the TCR design pattern without needing to write code at all will open the doors for the less technical users to leverage the technology.

The canonical TCR design consists of a Partial Lock Commit Reveal scheme. As the library of Aragon apps grows, organizations may shift away from the PLCR voting process. And instead opt for a delegative voting application to resolve concerns about the incentive mismatch between small stakeholders and TCR whales. Implement private voting to minimize the ability for curators to be influenced by bribery. Or replace voting entirely with prediction markets. Most of these TCRs will fail. But the cost of failure will be small. And with each failure the community and ecosystem will have one more data point to iterate and refine the DAO economy. As more DAO entities begin to thrive and create economic value, there will be greater need for a digital jurisdiction. A jurisdiction that caters to their unique needs.

Speaking of delegate voting...

Approaches to delegated voting is something that we have been exploring. Being able to delegate voting authority is a clear way to improve voter fatigue and apathy. There are many cases where we want people to have authority and would prefer if they voted directly. But we cannot expect them to be able to stay deeply informed. Or have the relevant expertise to vote on every issue by themselves. By delegating votes you have a smaller pool of people that need to vote for a decision to pass. Less people voting, but still keeping the final authority in the hands of a much larger and more decentralized pool. This helps so that decisions actually happen, and with less transaction cost overhead.

Another important, though slightly more nuanced advantage, of delegation is that it makes it much easier for DAOs to participate in other DAOs. They can delegate their voting privileges to individual members, or small well coordinated groups. This is preferable over a cumbersome and slow internal governance process to perform the desired action.

Some of the approaches to on-chain delegative voting we have been exploring are described below.

Liquid Democracy

Liquid Democracy allows creation of delegation chains based on topics. This creates a very dynamic structure where users have the authority to override their delegates. Or their delegate's delegates. It's as if you elect someone to represent your interests, but as soon as they do something you disagree with, you can pick someone else. Or vote on that issue yourself. Liquid Democracy offers a compelling alternative to Direct and Representative Democratic structures. Earlier this year we worked to create a set of requirements for on-chain liquid democracy. Yet, getting to a technical implementation that we are comfortable with, from a cost and security perspective, has proven difficult. Jorge did an excellent write up on the open challenge of implementing liquid democracy on Ethereum recently. Check it out and if you have any brilliant solutions please share them!

ERC900 and Staking Key Standardization

We are promoting a staking standard (ERC900). It would allow applications such as our voting app to be compatible with multiple staking implementations by design. However, ERC900 does not currently have any provisions for delegation. Which means that any action taken by the staker would need to come from the same key that used to stake to begin with. This is not ideal, as a user may not wish to use that same key for security reasons. Or the entity staking and un-staking may not be the same entity assigned with the responsibility of performing actions such as voting. An interesting use case for that type of organization might be a staking pool DAO. The technical implementation discussion on staking key hierarchies can be found at the research forum.

Council Voting

Another form of delegation that is of interest to us, is the idea of using an elected council as a supplement to direct voting. This is due to the challenges of efficient implementation of liquid democracy on-chain. This idea was inspired by the proposed mechanism for Polkadot Networks Governance. They are using a technocratic council to dynamically adjust the quorum and threshold of a direct vote. Those are dependant on whether the issue is contentious or not within the council. If the council is elected for fixed terms this looks more like a representative structure. But can be made more dynamic and closer to delegation by using an on-going election mechanism like a TCR.

This type of voting structure should be fairly simple to implement. Implementation can be done by using the basic building blocks we have created so far. We look forward to experimenting with this in the near term.

Deploying the first experiments

We are eager to deploy meaningful experiments to the Ethereum Mainnet. That's why we recently outlined a clear proposal for the first step. Using this structure we can allow ANT holders the right to veto any future funding proposals by the foundation. Giving them a meaningful and binding role in the projects governance. From there, we can propose further experimentation. This would come in a form of deploying DAOs that have different governance processes which would be proposed to ANT holders in order to fund them.


To keep up with the progress of Aragon:

Subscribe to the Aragon Project blog

Come chat with us at the Aragon Chat

Follow Aragon on Twitter

Subscribe to the Aragon subreddit

Follow Aragon at LinkedIn

Contribute to Aragon at GitHub

Find us on YouTube

Subscribe to Aragon Monthly Newsletter

Explore the Aragon Wiki

Browse job openings at Aragon

This post was a collaboration between

Luke Duncan, Aragon, Aragon One

  • Luke Duncan

    Luke Duncan

    Research at Aragon One

    More posts by Luke Duncan.

    Luke Duncan
  • Aragon

    Aragon

    Build unstoppable organizations on Ethereum. Allowing the creation of value without borders or intermediaries

    More posts by Aragon.

    Aragon
  • Aragon One

    Aragon One

    Aragon One is a for-profit company that encompasses the foundational team working on the Aragon Project

    More posts by Aragon One.

    Aragon One
Aragon Labs Research Update #1

Subscribe to Aragon Project Blog