'The Network Union' - The DAO State

A review and extrapolation of Balajis recent essay

Open Source Knowledge = Positive Sum

Siloed Knowledge = Zero Sum

Introduction:

This article is a response to a post by Balaji Srinivasan 'The Network Union' posted on the platform 1729.com

1729.com is an incredibly intriguing platform to add value to the global populace. Largely the premise is stepwise education by offering bounty tasks. Incentivized learning with a 'Github' of shared tasks and tool development. This allows global competition in pursuit of worthy knowledge and experience, which is in turn open sourced. The implications of this platform are massive. Global collaboration on difficult problems, relevant education and compensation for anyone with internet access, and potentially the mycelium for a powerful decentralized network state. This post is intended to continue the conversation regarding the latter.

---

The DAO Union State

The premise of ‘The Network Union’ is to create a functional organizational methodology for the operations of a decentralized network akin to a country. The union is similar to a typical workers union, its primary goal being to represent the interests of its members. Being the relevant governing agency for a decentralized groups allocation of resources, communications, and enacting its protocols is however no simple task. There are several questions that need to be asked, and processes developed to deal with them.

What are the pitfalls of modern day centralized governing bodies which a decentralized society intends to avoid?

• The law of averages - “there is no one size fits all”

An appropriate illustration of the struggles of many modern day governments is the law of averages. Elected officials have a recurring obligation to convince as many voters as possible they will enact the policies that are best for them, which indicates that by default they will fail.

An example is the H1B Visa program and immigration in the US. By all logical calculations immigration is a net positive for the US. High skilled workers contribute with high wages in fields where the labor supply is short, not only do they contribute directly by adding value to companies and being taxed on their wages, they also contribute to creating many new jobs associated with the products they work on. Low skilled workers also directly benefit the US by accepting lower wages and in turn driving the price of lower skill goods and services down.

Broadly this is a net positive, but it's also a fact that then the native low skilled workers are competing for jobs with lower wage expectations, likely now in a community where some high skilled peers are also earning more, they may be, or at least feel, worse off. Which leads them to demand their elected officials reduce immigration. The elected official panders to the populace by offering something of an average, limiting immigration, leaving the entire group slightly worse off instead - fitting no one. You could explain how immigration makes us all better off to the low skilled native worker but it’s a complex and highly obfuscated argument, which is very hard to feel at an individual level. Not to mention that much of the value add is captured in government bureaucracy and never returned to the individual.

In a decentralized governance this process can be handled by qualifying and appointing an ‘immigration’ management organization whose sole purpose is creating economic prosperity, which is experienced by every individual directly through the value in their digital wallet.

Some Immigrant Market Cap

• Authoritarianism and Appointments

One odd function seen in modern day governmental bodies is that of the appointment. An individual could rise to power in a strong man government, or be elected democratically, then they or a small body are solely responsible for appointing various officials tasked with increasingly broad swaths of policy enactment. For example in the United States you may have voted for a president because they seemed to be proponents of gay rights, voted for your senator because they offered to lower your state taxes, and then through some poorly formed logic they then become responsible for deciding who will manage agricultural policy for the entire United States, despite likely knowing next to nothing about agriculture themselves.

As the disparity between scientific comprehension of elected officials and the necessity of scientific capability required for the roles they appoint continues to widen the impacts can become devastating, which was on full display during the COVID-19 pandemic response.

These are two examples of how centralized hierarchies in government can fail to reach desired outcomes, but there are many more. The key to a successful decentralized union state will rely on the ability to retain democracy and individual choice, but to design hierarchical structures based on core competencies and the quality of their execution. Aspects of modern democracies were a good step in societal governance but were designed for an era that existed before instantaneous access to information and communication were possible. This ability has eroded trust in the modern institutions, but it also enables a way forward.

How will hierarchical bodies, subgroups, and individuals be validated so that their influence and power can be commensurate with the needs and goals of the entire group?

Social Clout

In 'The Network Union' Balaji highlights the existence of backlinks in modern social networks graphs and how they are currently used. Company founders and prestigious individuals gain social power and validity through the links in their existing online graphs, whether they be inherited or acquired in some fashion. Conceptually a similar link structure could be used to validate individual nodes (persons or entities) within the union network and ascribe them more influence or power for various tasks. This structure would be more akin to a tree than a graph, similar to the hierarchy of a large modern company, but with the goal of maintaining sovereignty instead of solely accumulating capital.

At face value this seems quite logical, but we will have to address some of the flaws in the existing system of ‘backlinks as proxy for validity’ that exist on the internet today.

The quality of a backlink is hard to quantify. There's an argument that some individuals rise to prominence (think politics) solely based on the number of links and subsequent impressions alone. Ability is hard to judge through back links. If Mike Maples Jr. and I are twitter mutuals it's far from an accurate indicator that I'm a competent and capable venture capitalist. Think of the Linkedin illusion, many people I know have 500+ connections on Linkedin, they may have even met some of them. Some of them may have vouched for skills listed, but whether or not they actually ever worked together utilizing those skills is not clear to be seen. SEO and social graphs have long been gamed by smart individuals to garner an audience for product sales. The mere existence of a large number of links isn't necessarily indicative of quality either, and can be manipulated algorithmically or financially. Putting the links on chain is a step in the right direction, but that merely indicates that they exist, not the quality of them.

In the Game

What this will require to be effective is skin in the game, similar to the concept of 'bitclout'. This could begin with crowdsourced open voting to set a baseline for individuals and then proceed like this: Imagine a video game character with varying skill values but instead it is sovereign decentralized state members. Coding level 47, Video Production level 84, Bioinformatics level 8, meme level 93, etc. In this game you can earn your skill level upgrades by accomplishing tasks, creating products, or have your skills upgraded by those who vote for you, but voting creates a dilemma. If you vouch for an entity and they fail, both parties lose points in that skill set. Reminiscent of the nodes of a neural network gaining or shrinking in their weight, this should optimize the functionality of the entire organism. Voters with higher or lower scores in particular areas will get higher or lower weighting to the votes they give. This can all be done computationally and stored on the chain.

Here is a potential example to put it into perspective. The Decentralized union state has a group of members representing agriculture in NE South America. The farmers are utilizing patented IP created by other members of the network to artificially extend their seasons and create profit in their local market, some of which is returned to the DAO. Another group of members of the union across the globe creates a varying light spectrum LED which could increase yield. The farmers choose to utilize the new light when a group of 12 members with botanist rank 80, and 1213 members with the ranking of 5 vote for the new light teams product skill set. If crop yields rise after the implementation of the product then the new members ranking improves and everyones value held in the DAO currency increases. if crop yields decline both the new team and all members who voted will have their skill levels drop at some ratio relevant to the reduction in yield. The majority of this happens computationally on chain in relatively real time. Optimizing for quantifiable functionality that is democratically enacted without the need to convene with bureaucratic intervention for or against as the procedures are largely all pre-determined.

How will decisions be prioritized, escalated, and acted upon?

The Paralysis of Choice

A country, digital or otherwise, has an immense amount of decisions to make at any given moment in time. Ideally a decentralized country has a truer to concept democratic functionality, yet the need for hierarchical structures remain, as it's required to segment decisions on various topics to the parties most affective at addressing them. It would be profoundly illogical for everyone in America to vote on the budget to fix the potholes on W Mill St. in Akron Ohio, and so too will many decisions relevant to a global distributed network be important but not demand attention from every member node.

Luckily the internet is already awash with methods to present massive amounts of information by relevance. Like the above mentioned validation protocol for member validation, a network tree vs. a social graph, so too can information be fed to the members for consideration in a prioritization tree. A database containing each possible issue to address can be created with any individual member adding issues they feel are relevant. A feed comprised of all of these issues will be made with members voting both on prioritization of the issue and if applicable their vote. A 'front page' feed with all of the 'trending' highest priority items can be made available. Each individual member will have their realm of interest or competency, by tagging and categorizing issues that are listed the network can insure that relevant items are initially displayed to the right members of the community. All continuously searchable, only the most pressing showcased.

As each individual will have ratings for their areas of expertise a vote for prioritization from a member ranked highly in a particular domain will add weight to the distribution of that item for other members also in that domain. This is not dissimilar to how social media functions, but distribution will be optimized for relevancy and priority vs. engagement metrics. Much of the prioritization ranking can be handled autonomously based on subject matter, and the threshold of the voting power needed to enact a particular motion can scale with its priority. The current ability for members to vote can fluctuate on or off with its prioritization level as well. This is not dissimilar to how some networks handle oracle management, inhibiting oracle functions if there is a perceived threat to the network. In this case small low priority items will often be handled autonomously through designed voting protocols embedded in the chain by relevant members, if there are contentions and prioritization is autonomously escalated to a certain point the ability for autonomous voting will be suspended and the motion will not be able to be enacted until there is a manual human voting of large enough proportion. Escalation can continue all the way up to the point of the entire chain needing to vote on an issue in the worst case scenarios (think ETH hard fork). This may sound complicated but with the current level of artificial intelligence it has been often shown that automated systems integrated with human oversight tend to outperform either alone.

Some Examples from medical imagery with ‘Human in the Loop’ methods -

https://www.nature.com/articles/s41746-019-0189-7

https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1002699

Here is a hypothetical example to better wrap your mind around the concept. A member of the network crowd funds the production of a digital art generation tool from the network. As its art is sold a portion of the sales are returned to the network. At some point the art generation tool creates a work that is perceived by a public facing company to be copyright infringement and a cease and desist letter is sent. A legal compliance node flags the letter and freezes transactions from the account associated with that art line.

The member generating the art claims it is not copyright infringement and lists the issue. It is flagged as legal and distributed to the most relevant legal compliance qualified member nodes with moderate prioritization ranking. Those nodes may be automated with legal compliance software analysis tools owned by network members. The tools review the documents and are weighting their analysis as low quality. The issue is flagged higher for prioritization. Human legal qualified nodes receive the issue within their feed and vote to a sufficient threshold that it is indeed copyright infringement. The member generating the art will now adjust their software and re list the issue as addressed with an existing prioritization level and any other qualified human legal compliance nodes will then be able to vote again. If the vote meets threshold then the artists account is unlocked again for transactions with the public.

At every touch point there are financial transactions based on prioritization and node member ranking. The artist could financially seed the prioritization to ensure it is resolved more rapidly, members could pay to add to the prioritization if they value the art or request a share of future revenue, legal nodes who want to engage in larger transactions could choose to ignore it until it hits a certain threshold and legal nodes who want to be more active can pick it up for less of a transaction fee. The network is safe from legal action, each actor is financially incentivized, the issue is scaled in proportion to relevance.

• — These some of my personal thoughts on the potential operational processes and possibilities of creating a decentralized network union, which will be one of the most important precursors to any sort of decentralized network country. This is by no means an exhaustive review but food for thought on design and functionality. There are more concerns such as computational demand and the costs incurred therein, network safety and attacks, and avoiding social/financial capital corruption. Perhaps the most important concern will be attracting quality members to the network who can help contribute to its design and adoption. If this is of interest to you please share the original essay from Balaji "The Network Union" and/or share this post.

  • You can also follow me on Twitter @BenjaminGCox or Subscribe to this email.