The concepts of DAO/community explored in this article will be referred to as “community collectives.” At the current stage, whether discussing DAOs or online and offline communities, they represent overlapping but substantively similar concepts. To better elucidate the commonalities between these two, this article will discuss DAOs/communities in a superimposed state. Additionally, the term “community” in this article includes offline communities.

Thus, whether you are discussing DAOs or communities, any exploration of the topic of “governance” falls within the framework of this discussion. On the historical timeline of technological development, from 2016 to 2023 marks the initial theoretical exploration and experimental period of DAOs. The new wave of artificial intelligence that began in 2023 has accelerated the advent of a human-machine symbiotic society, pushing DAOs and communities into a new development cycle.

In this new cycle, AI governance will take a leading role. AI models’ comprehensive capture of personal data will become commonplace. For example, Apple’s Personal Context technology captures extensive local data from iPhones to provide users with what the model considers optimal decision-making assistance.

Whether it’s governance decisions in DAOs, collective decisions in communities, or decision assistance from AI assistants, we are witnessing a significant societal transformation. This transformation affects the survival conditions of every individual and organization in a human-machine symbiotic society.

The granularity with which large models capture personal data will significantly enhance their reasoning abilities in user-specific scenarios. Furthermore, as various AI models are increasingly integrated into online systems to enhance tool intelligence, the competitive drive in technology will compel commercial companies to use all means necessary to obtain more personal privacy data.

Therefore, in an era where human governance trends toward AI governance, the extensive application of tool systems will unconsciously drive humans to quantify all individual behavior data metrics in DAOs and communities. This represents a machine’s KPI system acting upon human social activity systems.

Whether we like it or not, this trend is irreversible. However, we must be acutely aware in advance of what these quantification metrics mean for us, how the quantification of personal privacy data metrics interferes with our daily decision-making, and subsequently, how it affects our social cooperation relationships.

DAOs and communities embody our aspirations to break away from traditional cooperative organizations and seek egalitarian and fair cooperation. However, they will inevitably face new developmental challenges. Thus, this article uses the “governance paradox of quantifying public contributions” as a starting point to explore the fundamental contradictions in adopting quantifiable contribution governance mechanisms within DAOs and communities. It also examines how AI quantification fairness metrics, acting as a double-edged sword, create biased consensus and unfairness.

01 The current quantitative dilemma of community development

1. Common Questions and Deeper Issues

It is well-known that in DAO/community governance, seemingly egalitarian democratic voting systems can lead to power concentration within the DAO structure. Even with representative democracy, a few core members can monopolize decision-making and execution power. This is an inevitability in the classical structural model of DAOs, where decision-making and execution powers are intrinsically linked.

When decision-making power is concentrated in the hands of a few core members, participation in governance inevitably declines. These few members, from a game-theoretic perspective, hold the control and priority allocation of community public resources. This “power” relationship is not reflected in the “proposal-vote” action.

In fact, DAO/community governance structures exhibit uneven distribution of power relations, further leading to the democratic means of “proposal-vote” not genuinely granting individuals effective personal power. This results in a decreased willingness of non-core stakeholders to participate in governance. Differentiation among members inevitably leads to differentiated governance power.

Global DAO builders have now demystified the “democratic voting system.” Reflecting back, we misplaced our developmental will within the narrative framework of liberal capitalism, resulting in a collective illusion about true freedom and democracy.

Having traversed this detour, we are now able to reexamine our past experimental errors from historical and social perspectives. To overcome the governance dilemmas of DAOs, we must confront some fundamental issues, such as deconstructing individuality to build publicness, confusing the boundaries between community and publicness, token incentive mechanisms overshadowing the cultural order’s role in shaping organizations, and inequality in public property rights stifling individual development within DAOs.

We continue to face many problems today, requiring more researchers dedicated to both theory and practice to overcome our current challenges. The surface-level issues mask long-standing sociological dilemmas.

2. Governance Mechanisms for Quantifying Public Contribution Behavior

From the core issues of DAO/community governance, we can identify our fundamental demand for DAO/community governance: the pursuit of an optimal solution for the “fair distribution of public resources.” Therefore, we generally use the method of quantifying public contribution behavior to determine how public resources are allocated to the various members contributing to the community.

Token systems and point systems are common ways of quantifying the value of contribution behavior and converting it into cash (where cash here refers to a measurable unit of value; points/tokens are measurable units).

We attempt to define certain behaviors as having positive contribution value to the entire community. Thus, we use a points reward system to incentivize community members to actively engage in more contributive behaviors. Community members can convert points into cash/benefits. Points serve as a medium for realizing and trading contribution value, functioning similarly to currency.

For crypto communities, token incentives aim to address the same governance needs, but they focus more on using technical and monetary mediums. For instance, on-chain activity data is used as the valuation basis for token incentives.

Intuitively, we believe that quantifying contribution behaviors can establish an objectively fair economic reward mechanism. This mechanism allows us to clearly see each person’s contributions, thus achieving the fair distribution of public resources. This is the superficial reason why we generally introduce point statistics systems and token incentive systems.

3. The Curse of Quantifying Public Contribution Behavior

Adopting the quantitative governance methods of point systems or token incentive systems seems to be an inertia driven by our experiential understanding of socioeconomic systems. A good economic system can promote the prosperity and development of society. However, a careful examination of both ancient and modern times across different countries reveals that no economic system can perfectly solve the problem of fair social distribution.

Different economic systems have functioned at different times, but society is a more complex system, and economic systems always fail at some point. Sometimes, initially effective economic systems even exacerbate social wealth disparities, contradicting our original intention of seeking a good economic system.

The initial intention behind quantifying public contribution behavior is good, but reality often diverges from ideals.

When we attempt to construct an optimal solution for the “fair distribution of public resources” through quantifying public contribution behavior, in fact, precise numerical calculation systems also allow individuals to seek their personal maximum benefit and optimal solution within public resources based on quantitative indicators. Clear numerical indicators become excellent tools for benefit calculation. Since the rules allow it, we often only realize the severity of specific issues when individual profit-seeking behaviors disrupt the fairness boundary of public resources, but by then, it is often too late.

In the early stages, the points system incentivized contributive behavior and continued to create a spontaneously contributive atmosphere with subjective initiative. This atmosphere led individuals to spontaneously engage in various non-quantifiable, undefinable contributive actions.

When the non-profit-seeking subjective contribution vibe (a subtle “ambiguous” atmosphere of community values that allows non-utilitarian contributive behaviors to have influential power) is disrupted, those contributive behaviors driven by social and cultural value recognition will significantly diminish. Thus, profit-seeking behaviors under the rules destroy community fairness, and the systemic problems are difficult to resolve in the short term. This inevitably leads to the disappearance of many invisible contributions and the withdrawal of relevant personnel.

02 The Stacking Paradox of Quantifying Public Contribution Behavior

1.Our Intuitive Perception of Economic Incentives

In our common sense, it is intuitive to believe that when someone makes a contribution beneficial to the community, they should naturally receive economic rewards. This is almost an unquestionable consensus among all of us regarding this mechanism.

However, we should further examine the preconditions that lead to this intuitive understanding. I believe there are at least two reasons for this: one stems from our experiential understanding of socioeconomic systems, where labor results in deserved rewards; the other stems from our moral sense, shaped by our historical context and social culture, which ingrains in us a sense of fairness and justice—good people should be rewarded, especially those who contribute publicly.

It is our social experience and moral sense that give us this intuitive, albeit unexamined, recognition that incentivizing community contributions through quantification is feasible and reasonable.

This form of recognition of quantifying contributions involves a subjective interference with objectivity, leading us into the trap of experiential logic. Therefore, we easily encounter the paradox of something being “intuitively true but objectively false.”

The Stacking Paradox of Quantifying Specific Concepts

Regarding the governance mechanism of quantifying public contribution behavior, it actually consists of two forms: discourse form and measurement form. The discourse form interprets behavioral symbols, while the measurement form quantifies the degree of behavioral actions through quantitative research. In the measurement form, there are issues related to the boundaries and extent of action occurrence/execution. Therefore, we prioritize discussing the stacking paradox in the quantitative research aspect of the measurement form.

What is the stacking paradox?

The stacking paradox (Sorites paradox), also known as the heap paradox, involves a series of problems related to vague predicates and the accumulation of incremental changes. For instance, if one grain of sand is not a heap and adding a single grain of sand to something that is not a heap still doesn’t make it a heap, then no matter how many grains you add, you will never get a heap. This paradox highlights the issue of defining when quantitative changes lead to qualitative changes, which is directly relevant to our discussion of quantifying public contributions.

In the context of quantifying public contributions, we face similar challenges. Defining and measuring the exact value of contributions can be problematic, as small incremental contributions may not be recognized, but their cumulative effect is significant. This leads to difficulties in creating fair and effective incentive mechanisms that accurately reflect the true value of each individual’s contributions to the community.

What is the Sorites Paradox?

The Sorites Paradox, also known as the paradox of the heap, is a philosophical paradox that deals with the issues of conceptual boundaries and vagueness. The paradox can be illustrated through the following reasoning:

1. One grain of sand does not make a heap.

2. If N grains of sand do not make a heap, then N+1 grains of sand also do not make a heap.

3. By recursion, we can conclude that N+1, N+2, N+3, …, 1,000,000 grains of sand do not make a heap.

4. However, if 1,000,000 grains of sand do not make a heap, then adding one more grain should also not make a heap.

5. But following the recursive reasoning, we would conclude that 1 grain of sand makes a heap.

Thus, we find ourselves in a contradiction, unable to determine when a heap of sand transforms into a non-heap of sand and vice versa.

The core issue of the Sorites Paradox lies in the vagueness of conceptual boundaries and the continuity of change. It reveals that in certain cases, our conventional concepts and classification rules cannot be applied to boundary situations, making it impossible to determine when one state transitions into another. This paradox challenges our intuition about concepts and classification.

It implies the difficulty of conceptual classification because, during the recursive process, we cannot pinpoint where or when the transition occurs. This provokes thoughts on boundaries and vagueness, and questions the rationality of conceptual classification and definition.

——From ChatGPT

3. The Logic of Boundary Transformation Determined by Subjective Will

A natural extension of the Sorites Paradox is how we define the transformation of certain actions into public contributions. For example, in some community governance models, attending meetings earns points. In a community that values participation, any involvement in public activities is deemed worthy of incentive.

However, in a results-oriented society, merely attending meetings does not directly measure contribution value. Hence, simply participating in a meeting would not be incentivized. This logic represents our intuitive interpretation of contribution actions.

In a community that values participation, attending weekly, monthly, or quarterly meetings becomes an incentivizable contribution behavior. However, there is a difference between attending a meeting for one minute and attending for one hour. Since participants in a DAO/community can exit meetings at any time between one minute and one hour, how should we reasonably set the gradient of the reward scale?

Based on the time dimension, we further introduce the communication interaction dimension. Communication interaction is a deeper level of participation than merely attending a meeting. How do we measure the potential number of interactions, the number of interaction participants, and the relevance of interaction topics that could occur between one minute and one hour? This presents another challenge.

When we use quantitative forms to evaluate two contribution dimensions, the complexity increases significantly. If we adopt quantitative forms as the primary method for evaluating contributions, we inevitably push the system towards greater complexity.

As the system’s complexity increases, with the calculation of boundaries and continuous degrees becoming more demanding, the labor cost for community governance personnel also rises sharply. This can lead to a state of measurement redundancy and an unsustainable cost structure, ultimately trapping the entire system in a state of inefficiency and unmanageable overhead.

4. The Volatility of Subjective Value Boundaries in Open Communities

The collective subjective will that forms a consensus within a community is, in essence, a discourse-based consensus. This consensus is mainly achieved through interpretivism, which involves reinterpreting and reconstructing meanings. Interpretation is a deep description of symbols, and symbols are the medium through which we achieve consensus.

In a community, the open and fluid structure means that consensus is primarily attempted through “communication and interaction.” This is why many DAOs/communities, when faced with governance difficulties, seem to have endless meetings (debates/arguments/criticisms, with few in-depth constructive discussions).

However, the open and fluid personnel structure also leads to the collective subjective will being in a state of flux, causing the baseline of collective decision-making logic to be volatile. The logic of interpretation is constantly changing. Although interpretive logic deeply influences the quantitative aspect, the surface of the quantitative form does not change significantly; it may only involve adding new categories to the calculation methods.

Thus, an open and fluid interpretive interaction structure ensures that the community’s value preferences for public contributions are not static. Time is a key factor in this consideration. For DAOs/communities, as structural models within social relationships, achieving continuity must account for temporal considerations.

“Any real historical sequence is necessarily complex in its temporality because it is a specific combination of different social processes with different temporalities. And any particular historical sequence may combine an overabundance of trends, routines, and events,” William H. Sewell Jr.’s analysis highlights the complexity of temporality in historical sequences. In sociology, historical sequences can be understood as time sequences, which are the basic narrative forms used to describe and analyze social phenomena.

It is essential to understand what is meant by “trends, routines, and events”:

• Trends are directional changes in social relationships. Historians often use terms like “rise” and “decline” to mark such temporalities.
• Routines refer to relatively fixed and repetitive activities, such as a stable and continuously developing activity pattern under institutional constraints.
• Events are a series of actions that transform structures, concentrated in time, capable of establishing new routines to change old ones, thus accelerating, reversing, or repositioning trends.

This temporal analysis model comes from William H. Sewell Jr.’s study of how a series of economic, political, and technological factors in different social contexts changed the decision-making basis and value orientation of dockworker communities. This is exactly what DAOs/communities currently experience in their development.

For example, during the peak of a crypto bull market and the period of blind faith in democratic voting systems, community contributors were optimistic about the future and willing to pledge their contributions for token rewards and voting rights, seeking greater future returns. Conversely, during a prolonged crypto bear market and the disillusionment with democratic voting systems, community contributors, driven by pessimistic expectations for the future, refused to contribute without immediate returns and emphasized cash flow to ensure their contributions were duly rewarded.

This case illustrates how economic and political factors, as trends, change our routine behavior patterns.

5. Collaborative Strategies in the Game of Interaction Structures

Under the influence of temporality, the continuously shifting value preferences and fluctuating decision baselines in DAOs/communities inevitably lead to instability in the consensus interaction structure of the community. In such an unstable consensus interaction structure, community contributors are compelled to frequently adjust their collaborative strategies with the community, as their identities, positions, and value inclinations are easily swayed by the community’s consensus structure.

A community’s collective effort to safeguard public interests is built on establishing a long-term mutually beneficial relationship between individual development and community development through the consensus interaction structure. However, an unstable or even chaotic consensus interaction structure loosens and confuses this mutually beneficial relationship, ultimately leading to its dissolution.

In such scenarios, the basic stance of community contributors may shift from an altruism-prioritized mutual benefit relationship to a self-interest-prioritized interaction relationship.

6. The Hare Hunting Game: Abandoning Collective Interest Maximization

The principles of mutual cooperation and mutual benefit within a community rely on a stable consensus interaction structure. Once individuals lose trust in the collective mutual benefit relationship, DAOs/communities inevitably shift from pursuing a collective interest maximization model (Stag Hunt) to ensuring individual interest priority (Hare Hunting).

The idea of the Stag Hunt originates from Rousseau’s “Discourse on the Origin and Basis of Inequality Among Men.” The Stag Hunt describes a scenario where hunters can independently hunt hares to secure their basic survival needs. However, hunting a stag yields greater rewards, with returns far exceeding those of hunting hares.

Nevertheless, an individual cannot hunt a stag alone and must cooperate with other hunters. The more hunters involved, the higher the success rate of hunting a stag. If a hunter, during the stag hunt, spots a hare and opts to hunt it instead, it increases the likelihood of the stag hunt failing. Thus, hare hunting versus stag hunting becomes a game of individual versus collective interests.

In DAO/community governance mechanisms, the interaction form of the Stag Hunt should be our primary consideration. In reality, however, we often see various game-theory-related disputes in DAO/community governance discussions. Typical examples include the free-rider problem and the public goods dilemma.

The lack of clear collaborative strategies and positions of interest among participants in an interaction structure of mutual benefit leads to difficulties in understanding how specific public interest disputes arise and are resolved. Furthermore, it complicates our ability to determine which public games fall within a reasonable scope of definition. This is undoubtedly a challenging research task that requires significant investment.

Thus, when faced with issues of public interest, the DAO/community must establish a robust and reliable consensus interaction structure to encourage participants to prioritize collective benefits over individual gains. This involves creating an environment where the benefits of cooperation (hunting the stag) outweigh the temptation of immediate individual rewards (hunting the hare), fostering trust and long-term commitment to collective goals.

03 Labor exploitation and value alienation of invisible contributions by communities

1. Exploitation of Invisible Labor in DAOs/Communities

As previously mentioned, what constitutes a contributive action is defined by the interpretive framework of collective consensus, meaning the overall value preference of contributions reflects the collective will of the community. However, the consensus formed by weaker groups within the community often cannot influence the overall value preference of the community.

This brings us to the struggle for rights between feminism and capitalism. A housewife, for example, contributes significantly to managing the household, doing chores, and caring for the elderly and children. It is precisely through her labor that men can have reliable support in social production. From a sociological perspective, we cannot ignore the value women contribute to social and economic development.

However, in the logic of capitalism, housework performed by women is not recognized by the market and cannot be exchanged for compensation. The capitalist market system directly ignores the professional value of this labor identity, resulting in the ruthless exploitation of women’s invisible labor within the socioeconomic structure.

Similarly, in DAOs/communities, there are numerous contributive actions that cannot be collectively interpreted and measured. The exploitation of invisible contributions exists within DAOs/communities. Despite being aware that some contributive actions cannot be recognized in the short term, measures like contribution tracking incentives, welfare subsidies, and even self-empowerment (active claims for contributive rights) can be taken. Remedial measures can be implemented according to the specific conditions of the community, but they cannot cover up the fundamental and substantial problems.

The essential problem with non-quantifiable invisible contributions is the lack of collective interpretation (weak consensus) and measurement (no pricing). The consensus of dominant groups has blind spots in value preferences. This leads to the fundamental problem where contributions that are not collectively interpreted or have no discourse form cannot enter the structure of quantitative contribution reproduction, thus denying the reproductive value of non-quantifiable contributions from the production structure.

For a community, many spontaneous contributions that are not interpreted or measured by consensus, such as emotional value and intellectual value, constitute the abstract cultural symbol reproduction structure of “community-emotion-connection.” These essential elements are invaluable to the community, representing significant micro, diverse, and large-scale productive factors.

2.How Monetary Transaction Mediums Alienate Community Contributions

For a DAO/community, collective contributions should be diverse and spontaneous. Our recognition of public contributions is essentially an acknowledgment and respect for diverse values. However, quantification inevitably transforms the value of contributions into a singular monetary value because quantitative values serve as a monetary medium that must ultimately be converted into cash.

Contribution value is interpreted as a monetary unit’s measurable value, and the value of these monetary units corresponds to the consumer value of goods. Quantified contributions, through the medium of money, enter the commodity market’s trading system. Contributions in DAOs/communities, facilitated by monetary mediums, circulate within a broad economic market.

While this process helps move contributions from closed communities to open and expansive markets, allowing community contributors to gain higher returns in the trading market, it also transforms the community’s value logic of public contributions into the logic of commodity transactions in the public market.

When the mutual benefit relationship in the community’s interaction structure turns into a transactional relationship, for instance, when contributions are made to gain market funds or commodities rather than to consider the community’s sustainable development and value preservation, a fundamental shift occurs.

As self-interested profit-seeking strategies become prevalent in the interaction structure, capital transforms the structure into one aimed at maximizing capital reproduction. Capital captures the community’s reproductive structure and, through symbolic production, alienates the value concept of contributive labor.

This alienation occurs because monetary incentives shift the focus from communal values and collective goals to individual gains and market-driven transactions. As a result, the intrinsic motivations for contributing to the community’s sustainability and shared ideals are undermined, replaced by the extrinsic motivations of financial reward and personal profit. This shift fundamentally changes the nature of community contributions, eroding the social fabric that holds the community together and transforming cooperative efforts into market-driven exchanges.

3. Monetary Incentive Inflation Leading to Contribution Deflation

Monetary incentives represent an imbalanced economic model. To promote more contributive behaviors within a community, choosing a point/token incentive system inherently involves adopting a risky monetary policy. This policy converts a large amount of non-redeemable contribution value into monetary value.

The aggressive implementation of this risk-averse monetary policy continually leads to the inflation of contribution currency and the dilution of community contribution value. In such a risk-heavy monetary policy, the ongoing inflation of the currency results in the persistent dilution of contribution value.

A community’s development relies on business growth to drive effective economic behaviors. In the community governance mechanism, prioritizing a points-based system as an incentive method inevitably involves various approaches to issuing points/tokens to stimulate more contributive actions. This creates a seemingly logical growth model of “goal-task-currency-contribution.”

However, the points system as a monetary incentive not only serves the function of value transfer but also the critical function of value realization. Implementing a points system without establishing a sustainable development business is akin to injecting a growth stimulant into the community. The short-term prosperity it brings accelerates the community’s decline, which is true for any economy.

Excessive contribution output and currency hoarding, followed by insufficient contribution output and continued currency issuance to stimulate it, create an inescapable cycle. Governance mechanisms unable to break free from this cycle inevitably lead to the dilution of contribution value and the continuous devaluation of contribution currency. When currency inflation and value dilution occur, a community’s healthy contribution atmosphere will unavoidably be damaged, resulting in contribution behavior deflation.

In essence, as the community issues more points/tokens without corresponding valuable contributions, the real worth of each point/token decreases. This depreciation demotivates contributors, as their efforts yield diminishing returns. Consequently, fewer members will be inclined to participate actively, leading to a reduction in overall contribution levels, a phenomenon known as contribution deflation. Thus, the community must carefully balance monetary incentives to maintain the value and motivation for contributions, ensuring sustainable growth and engagement.

Last

Risks of AI-Measured Complex Governance Systems

Quantitative research in measurement forms is highly formalistic, while “contribution” is an interpretation of cultural symbols. We attempt to quantify an interpretive social symbol network system, which encompasses political, economic, and cultural elements—far beyond what we understand as a measurable contribution system from an economic perspective.

Quantifying complex systems is alluring yet extremely dangerous. It implies an attempt by public power to control an ultra-complex system while ignoring its inherent development laws. As measurement forms become increasingly complex, dealing with the intricate human interest relationships within public social systems becomes overwhelming, inevitably leading to calculation failures. This results in a series of measurement form breakdowns, culminating in the collapse of the public system.

As governance systems become more complex, humanity will inevitably turn to AI for governance assistance. In an era of human-AI symbiosis, humans will be unable to accurately judge governance conditions in specific scenarios and will likely delegate these tasks to AI. This is similar to the emergence effect of large language models, where researchers still don’t fully understand the principles behind intelligent emergence.

The ultimate goal of community governance is to achieve moral justice. Quantification is a means to measure the contribution value of community members and distribute resources fairly according to this value system.

However, as the governance procedures for quantifying public contributions evolve into a large and complex system, humans will inevitably introduce AI to assist in governance tasks. Humans will be unable to accurately judge specific governance conditions, and these tasks will likely be handed over to AI. Just as with the emergence effect of large language models, researchers still do not fully understand the principles behind intelligent emergence.

AI training data may contain untreated risky data, such as racial discrimination remarks, gender opposition remarks, and violent behavior data, leading to biases in AI’s understanding of moral justice and causing governance crises in specific situations.

Ensuring AI consistently makes correct decisions in a complex human governance environment is challenging. Diversity in training data and the construction of a distributed governance system theoretically help AI make more objective and fair decisions. However, in an anonymous decentralized governance system, witch attacks can be launched using multiple anonymous accounts to initiate Proof of Unlearning attacks, deleting specific training datasets from the model. Alternatively, injecting polluted data into distributed training models can cause bias in the model’s predictions. This is a form of reverse interference attack on the attention mechanism.

Most current research on AI governance remains in the academic field. However, with rapid technological advancements and humanity’s increasing reliance on digital governance systems, we are bound to face a more complex governance environment.

Disclaimer：

1. This article is reprinted from [VION WILLIAMS]. All copyrights belong to the original author [VION WILLIAM]. If there are objections to this reprint, please contact the Gate Learn team, and they will handle it promptly.
2. Liability Disclaimer: The views and opinions expressed in this article are solely those of the author and do not constitute any investment advice.
3. Translations of the article into other languages are done by the Gate Learn team. Unless mentioned, copying, distributing, or plagiarizing the translated articles is prohibited.