Galactica Network's Proto-Governance Model
Introduction
Similar to how USA’s political system has evolved from the Congress-only state of provisional government of Second Continental Congress to a massively sophisticated system we know today, Galactica Network’s Cypher State will be deployed in stages, persistently growing in sophistication as the number of Citizens grows and the Network matures.
This document outlines the way the Galactica Network Governance will be implemented before its final version, which is explained in:
a) Galactica’s Academy of Science and Governance of Ecosystem Funding [1], and
b) Galactica Network Governance Framework [2].
Note, that while reading these documentation is not strictly required in order to understand the Proto-Governance model, we highly recommend familiarizing yourself with them before proceeding.
Campaigns, SBTs, and GNCs
Developing a Governance model that begins with the TestNet is of practical importance, since we believe that early user participation is strictly necessary to create a short list of the Galactica’s first Citizens while the technology itself is being tested and audited before the MainNet launch.
Users earn GNCs, while taking part in activities during the DevNet and the TestNet, while the GNCs are activated as of the start of the MainNet, therefore there are no practical implications of imposing a specific Reputation function on the TestNet. User participation is quantified with the use of SBTs that are campaign-specific as campaigns directly benefit the ecosystem. Involvement and impact is noted in the campaign process and users are rewarded directly in the form of additional Voting Power that depends on the SBT type. These SBTs will be transcribed into initial Reputation scores once MainNet is launched.
Timeline
During the TestNet the Governance model will pass through the following stages:
TestNet Proto-Governance I
TestNet Proto-Governance II
MainNet Proto-Governance
MainNet Fully Fledged Governance
In brief, the TestNet Proto-Governance will begin with a few entities that facilitate the formation of communities focused on the development of the Governance process of Galactica and eventually become active participants in the Governance. The Proto-Governance of the MainNet incorporates elements of decentralized funding for projects and eventually evolves into the fully fledged Governance which is described in the two documents mentioned earlier.
As a reminder — to participate in the governance of Galactica, users are required to join Interest Groups. This is similar to voters joining a political party, or Swiss citizens in a Canton. The Galactica governance structure comprises the Parliament, which consists of the National Council, the Council of Interest Groups, and the High Council. Galactica’s system of governance combines characteristics of modern day democracies with a meritocratic focus. By incorporating merit into Galactica’s governance framework, we aim to address common challenges faced by traditional democracies and provide users with an equal opportunity to contribute and be duly recognized for their efforts.
TestNet Proto-Governance I
Entities & Roles
At this point relatively few users will comprise the Galactica community — creating the Councils with sufficient number of Representatives is not a viable practice as it will increase centralization of power later on as these people must to be rewarded with Reputation. Therefore, the following process is proposed:
Initially, a platform named the Forum will be implemented. Its purpose is to introduce a playground where users can exchange and discuss ideas relevant to the network.
Proposals are chosen centrally by the Foundation at start based on proposals submitted on the Forum.
Proposals are created by the community.
Users who have submitted proposals that are approved by the Foundation to be voted on will eventually be rewarded with Reputation (thus, UBI at the inception of the MainNet).
TestNet Voting Power (VP) Definition
The VP function is calculated for each user separately and the function is specified as follows:
where,
VP tokens — token-based portion of the VP function. Tokens are obtained via a Faucet thus their impact needs will be constrained to a predetermined interval, otherwise users will be able to claim amounts that will make their VP substantially larger in comparison to other users.
VP Campaign-i — Reputational part of the VP function. Galactica’s campaigns (e.g. the Cypher State campaign, Voyager Ark campaign, Sentinels) will reward their users with Soul Bound Tokens (SBTs) — non-tradable representation of one’s merit — which are going to be used in calculation of one’s VP thus introducing merit-based voting.
VP Additional SBT-i — Additional SBT rewards that factor into the VP. These are awarded by the Foundation for extraordinary contributions.
Token Component of the VP Function
As noted in the paragraph above, Faucet tokens are free and having their impact on the VP unbounded will dramatically influence the incentives of the system. The token component is determined as a function of the number of tokens (x) obtained via a Faucet in the following way:
where,
p — rescaling parameter. Determines the impact a token has in the VP function.
b — determines the function shape.
For more details see the following visualization example: Desmos
The Merit Component of the VP Function
As specified in the Galactica Network Governance Framework paper, eventually Reputation will quantify a user’s impact in the development of Galactica Network. During the TestNet the merit will be directly incorporated into the VP function. Campaigns are built in steps (Tiers), where each has a predetermined Difficulty Score. This score will, in theory, correspond to the amount of work one needs to put into the Network. As the difficulty score represents work (weighted by quality) it will serve as a base for the simplified Reputation mechanism.
The Merit-based portion of the VP is a function dependent on these Difficulty Scores and, in principle, follows achievement difficulty that is additive across the various Campaigns. It is defined across three different sections of difficulty (that may be changed with a Governance vote):
Superlinear Section — this section is designed to provide moderate incentives.
Linear Section — true meritocratic behavior as users’ VP is directly proportional to the amount of quality effort made.
Sub-Linear Section — reserved for high accumulated difficulty, with few users and therefore sub-linear behavior with a horizontal asymptote is implemented.
For the TestNet a simplified version of Reputation has to be developed that incorporates an upper bound. Not all users have equal capabilities (and same starting points in practice) — in order to prevent unbounded differences between them an upper bound is imposed. This solution needs to be implemented at the beginning in order to prevent the centralization of power. The bound can be changed by popular vote, and, since there will be only a few users in this section, there is no room for the creation of a self-reinforcing cycle that will place the system in a state of a dictatorship. If the system is left unchecked and the linear part continues to infinity, users will be effectively segregated by individual capabilities. Capable ones will, given enough time, rule the network. This mechanism design will maximize the wellness of a capable few at the cost of others, thus the function that defines their Merit-based portion of VP must be parametrized through the voting process itself. Eventually, the system will settle at a function that both prevents dictatorship by the capable and also incentivizes them.
From the mathematical standpoint the function is to be differentiable in order to avoid sudden jumps in VP and keep the incentives continuous between sections. The function is defined on 3 intervals (sections), that are:
Section 1: (0, a]
Section 2: [a, b)
Section 3: [b, + )
Section 1
where,
α — slope of the linear part
β — shape parameter
Section 2
Section 3
where,
t — a parameter that determines how fast superlinear part converges to a horizontal asymptote;
γ — shape parameter;
Note: Parameter is first to be defined and for each value there exist a uniquetsuch that the function is tangent to the line from the linear part.
The system can be configured with 6 parameters:
a — end of first interval;
b — end of second interval;
α — linear segment slope;
β — first segment shape factor;
t — defines the scope of the sublinear part, maximum VP;
γ — shape parameter.
For more details see the following visualization example: Desmos
Discussion
In this context, one can define meritocracy “rewards and benefits are to be proportional to the amount of effort that has been put into the system.” Even if one finds an adequate function to measure the quality and quantity of effort that has been exerted the definition of a fair system is still problematic.
A hypothetical function that translates all different kinds of effort is assumed (though its exact form is likely next to impossible to find), and it will give each user a difficulty score in practice. These numbers are to be translated into Reputation in order to introduce a connection between effort and the corresponding reward. If one doubles the effort made, the reward (Reputation) should double as well — quality effort is translated into Reputation.
This line of reasoning implies a linear dependance. Other than being meritocratic, this system also becomes heavily impacted by the user’s individual capabilities. Without any balancing mechanism, this system will progress into a system ridden with inequalities, and a community of elites may be established. One inherent balancing mechanism can always be introduced — Reputation that is bounded by some predetermined constant, thus the function will possess a horizontal asymptote. Furthermore, participating in the governance of protocols in the digital assets space has exhibited relatively low interest/participation rates in the past. A relatively simple incentive has to be introduced at the beginning as the entire system must be kick-started. One such incentive would be found in the super-linear portion of the Reputation function, as the rewards would increase faster than the effort made for the benefit of the system.
If the function was super-linear, throughout its entire domain, receiving rewards would be getting easier and easier given a constant input of effort. Examples corresponding with such behavior can be found within highly capitalistic systems. Inequalities are accumulated fast in these kinds of systems since they are heavily dependent on users’ capabilities but also their starting points. The sub-linear function is problematic since users’ reward returns get progressively smaller and smaller over time. This creates a negative incentive on the population as a whole to participate. Meritocratic system would be a linear one — a system that is between these two paradigms.
Anyone that tries to create any system that has to be fair will soon be humbled by the complexity of the problem. This implies that strict parameterization is not a good option; parameters need to be fluid and easy to be changed. In Galactica’s system, they can always be changed by the Voting process, and the people of the network will be able to choose the system in which they wish to participate. Cautious readers can find a problem with the circularity of this solution. Referendums are, for this reason, introduced later. Voting is done 1-person-1-vote, thus at the ground level the system is brought to democracy if an elite with a significant amount of Reputation and VP is created.
From the practical perspective, these paradigms are de facto different in the values of the parameters of the Reputational function. One of the most important parameters is the slope parameter seen in the first two intervals.
The slope determines the rate at which a user will receive their rewards per unit of effort. Its slope is defined on a system scale, therefore little difference will be made globally as a large part of the user base is expected to be in this section and slope change rescales the entire system. Therefore, users with a low and very high amount of accumulated effort done will be affected.
Shape factors both in super-linear and sub-linear parts only factor in the intensity of their non-linear characteristics. Parameters of greatest importance are a & b — they determine the end of the super-linear section and start of the sub-linear section. Differences in capitalistic / meritocratic / socialist behaviors are represented in the interplay of these parameters. As a good rule of thumb a should be small and b should be relatively large so that a larger part of users fall in the meritocratic region, however it can be changed through the governance process.
On the final note, these paragraphs aimed to convey the idea of a Reputational function. Reputation itself won’t be introduced at the beginning, users’ merit will be determined and they will be rewarded with SBTs representing effort excerted. These SBTs factor in the VP function and the merit plays this simplified role in the whole system. Reputation, once introduced, will serve a more significant role in the network. It will be utilized for UBI distribution, Contingent Transactions, VP and more.
Consensus
Certain minimum required VP (in % of the total) is necessary for a Quorum to be created. If this is not reached, an insufficient number of people participated (or insufficient amount of VP had been used) and the vote is postponed. There is a limit on the number of times a vote can get postponed, after it is reached it is marked as irrelevant and discarded. Let us name this percent X.
In order for the vote to be accepted / declined more than Y% of voting power needs to be allocated to the YES / NO option. If said condition is not met the vote is postponed for further discussion on the forum. There is a limit on the number of times a vote can get postponed, after it is reached it is marked as irrelevant and discarded.
Y is a function of X and for simplicity it can be seen as a linear function, since there must exist a cutoff we define X0 as a percent under which Quorum cannot be reached (e.g. 20%). If all people participated then X=100% and this point we define as Y=50% or in simple terms if all people vote consensus is at >50% of VP. Consensus for X0 is a parameter that defines the slope and we can name it Y0 but it must be strictly larger than 50%. Consensus is reached if votes allocated are greater than:
TestNet Proto-Governance II
Entities, Roles, and Election
In this phase a simplified version of The High Council (The Committee) is created and it will be the precursor of this governmental body. Their obligation is to observe The Forum and distinguish relevant from irrelevant proposals based on community engagement but also based on their personal assessment. They are the ones that accept proposals and should act in the best interest of The Network and its users.
Users within the Committee are rewarded with additional Reputation (thus UBI) if their proposals are voted on. Whether a proposal is accepted has no consequence on their rewards — if a Quorum is reached, they are rewarded. Mandates exist for representatives — their duration will be specified at a later stage.
People from the set of SIG Member Candidates (SIG MC) — future members of Special Interest Groups — are chosen within their Interest Groups either centrally, through VP, or by a Popular vote.
Committee members will be chosen form:
SIG MCs — 2 representatives are chosen from each SIG Member Candidates’s set, thus the total number of representatives is 2N where N is the number of SIGs.
Citizenry — ordinary users that earned or bought GNC and will get their GNC activated at the beginning of the MainNet. Three criteria are applied here:
a) Voting Power
b) $GNET stake
c) Reputation
Consider the following example:
4 SIG MC sets;
This implies 8 SIG MC set representatives;
Number of Citizenry-sourced is divisible by 3, therefore the amount that is maximum and still lower is 6;
2 representatives are sourced from each criteria (from ordinary users).
Representatives that are sourced from the Citizenry can be removed from the Committee by mandatory referendum rules, as specified in the Galactica Network Governance Framework document. After a certain number of people (measured in percent of total number of citizens) specify that the referendum must be held, a mandatory referendum will take place. At least 50% of users must vote on the options given and all of them have exactly 1 VP within this process (1-person-1-vote). If a Quorum is reached 50% for ordinary and ⅔ for Governance votes need to be collected for a single option for acceptance/rejection to occur.
The Committee serves the purpose of choosing proposals, which will be presented to the community and they need to reach a consensus in order for the proposal to be accepted. The same functionality is applied to this Committee as it envisioned in TestNet Proto-Governance I — the threshold of proposal acceptance is a function of the Quorum. Minimum Quorum threshold is at least 1 user per every SIG MC set and 1 per each category from the Citizenry.
MainNet Proto-Governance
Entities
Funding of projects during TestNet Proto-Governance I & II is done centrally via the Foundation. MainNet Fully Fledged Governance is the full governmental model as specified in the relevant documents. Important takeaway is that MainNet Fully Fledged Governance introduced decentralization of Academy of Sciences and other entities related to funding. Between a central version of these entities (TestNet Proto-Governance I & II) and its decentralized one (MainNet Fully Fledged Governance) a transitory model has to be established.
The MainNet Proto-Governance introduces partially-centralized funding processes.
The Foundation remains and gives out Grants (centralized) directly, however a portion of Ecosystem Funding is done in a decentralized manner through the Academy of Sciences (AOS) and later Parliament. These entities are composed of individuals from the community, which possess the relevant skills to carry out their duties and are remunerated for that.
The entities that will be firstly established in this stage are AoS and the Auditing Commission.
The Academy of Sciences would control the funding across different topics (development directions) and the projects for each topic.
Once the system is sufficiently developed certain obligations would fall over to the Sovereign and the Excelsior (see the Governance of Ecosystem Funding article for more details), however at the start of this stage they will be inside either the AoS, the Foundation, or the Auditing Commission.
Roles
Initially project onboarding, distributing funding across the projects in each topic group, and project audits are the only relevant tasks that have to be performed.
Funding is still centralized via the Foundation, however departments within are created for the specific tasks aforementioned. These departments function as a unit within the Foundation with a defined hierarchy. They are paid in $GNET for their efforts and are overseen by the Foundation, which has the right to select and dismiss these responsible individuals.
The Auditing Commission performs a holistic DD process on the projects applying for funding. Reports are written and sent to the AoS and the Foundation.
AoS reads the Audits given by the commission, discusses with the Projects on their KPIs and, overall, performs Project Incubation related to the Galactica ecosystem. AoS also searches for additional projects to launch at Galactica if too few have applied.
Everything is overseen by the Foundation that, at the end, gives the Grants, however note that Foundation centralized funding still exists.
The Sovereign is formed, however it is not yet operational. Funds that are going to be used at a later stage once full governance is implemented are simply being accumulated at this point.
Once the MainNet Fully Fledged Governance begins these units (AoS and the Auditing Commission) are removed from Foundation’s jurisdiction and placed into the jurisdiction of the Parliament thus ensuring decentralization, however with a set of capable individuals that perform necessary tasks within each entity.
Website | Twitter | Telegram | Discord | News | Reddit | YouTube | Zealy| Notion | CypherState | Telegram Community