Chapter 2: The Problem

Draft v0.1 - 2026-05-24 - awaiting Zaal review

Modern DAO Governance Does Not Work

In 2020, Compound launched governance. The idea was simple: let token holders vote on protocol parameters. Every holder with 25,000 COMP could propose. Every voter could delegate to a representative. This was hailed as a breakthrough - decentralized governance at scale.

Six years later, Compound has 8 delegates controlling 50% of voting power. Eight people decide the future of a $5 billion protocol.

This is not a Compound problem. This is a token-weighted voting problem. It is the default outcome when you link voting power to capital.

Uniswap has 11 delegates holding 50%+. Aave has concentrated delegation. Every major DAO using token voting has experienced the same curve: power concentrates. The Nakamoto coefficient - the minimum number of participants needed for 51% control - is 8 for Compound, 11 for Uniswap, 18 for ENS. This is not decentralization. This is oligarchy with a blockchain backend.

The second failure is voter apathy. In Compound, 5% of token supply participates in a typical vote. Uniswap averages 5-7%. Across DAOs, participation is 3-10%. This means 90%+ of token holders do not care enough to vote. The rational ignorance problem is alive and well: in a population of millions of token holders, my vote has negligible impact, so why spend energy learning about the proposal? I will not change the outcome. I will skip it.

Result: governance is controlled by a small group of institutional voters (funds, founders, early investors) who have skin in the game. Retail token holders are theoretically included but practically powerless. This is not consensus. This is theater.

The third failure is vote-buying and plutocracy incentive-design. When voting power is liquid - when you can buy more tokens and gain more votes - governance becomes an auction. If you want a proposal to pass, buy tokens. If you want it to fail, lobby the current big holders. This inverts the incentive structure. You are not rewarded for good governance; you are rewarded for capital accumulation.

The Math: Concentration in Token-Weighted Systems

Let us be specific.

Compound (as of 2024):

• Top 8 delegates: 50%+ of voting power
• Top 50 delegatees: 99.23% of voting power
• Polychain Capital alone: 11.8% of all COMP
• Participation rate: ~5% of token supply votes on typical proposal

Uniswap (as of 2024):

• Top 11 delegates: 50%+ of voting power
• Top 50 delegatees: 94.73% of voting power
• Participation rate: ~5-7% per vote

ENS (as of 2024):

• Top 18 delegates: 50%+ of voting power
• Slightly more distributed than Compound or Uniswap, but still oligarchic

These are not edge cases. These are the largest, most “legitimate” DAOs. If Compound - which was specifically designed to decentralize governance - ends up with 8 entities controlling 50%, the problem is structural.

The cause is simple: capital is already concentrated. Pareto principle is inevitable. In any population, 80% of wealth is held by 20% of people. Of those 20%, 80% is held by 20% of them (4% of the total population). Continue this cascade three times, and 0.04% of the population controls ~51%.

This is not evil. It is mathematics. Capital accumulation compounds. In token-weighted voting, you do not fix this; you amplify it. Voting power becomes a pure function of capital, and capital is already skewed toward early investors, founders, institutions, and the wealthy. Voting codifies this skew.

The data is clear: token-weighted voting has not produced a single large DAO that is meaningfully decentralized by any measure of power distribution.

The Deeper Failure: Capital Is Not Contribution

This is the mistake that Web3 governance has never solved.

Token voting assumes: people who bought the most tokens are the people most qualified to guide the protocol. This is sometimes true for price speculation, never true for governance.

A person who bought 1,000 tokens on day one at $1 each and held them to $1,000 each has not necessarily contributed anything to the protocol. They speculated correctly. A person who contributed architectural insight, code, community building, or partnerships may have never accumulated tokens early enough to accumulate significant power.

In a music community, this problem is acute.

A musician who released a breakthrough album that defined the community’s direction is not necessarily the person who accumulated the most tokens early. A community builder who onboarded 50 new members is not the one with the largest wallet. A mentor who shaped the artistic development of emerging artists may be less wealthy than a speculator who bought the DAO’s token on the secondary market.

Token voting inverts this. It says: whoever is richest makes the decisions.

In a music community, this is a betrayal of the community’s stated mission. If you are a music DAO, you are supposed to elevate music. Yet your governance system elevates capital. This is not a bug; it is a fundamental misalignment.

Music DAOs and Token Voting: The Case Study

Every major music DAO uses token-weighted voting because there is no alternative infrastructure.

Friends With Benefits (FWB): Token-weighted voting on Snapshot. FWB token holders vote on events, treasury allocation, and partnerships. Founders hold significant allocations. Early investors hold significant allocations. Recent community members - even if they are more talented musicians - have less vote power than early speculators.

SongCamp: Token-weighted voting. Matthew Chaim (founder) and early artists hold the majority. The community tested a “Chaos Agreement” (1-wallet-1-vote) once, approved unanimously. No sustained alternative emerged. They returned to token voting.

Catalog (formerly Audius-adjacent): Token-weighted governance on Creator (CRE) token. Catalog was supposed to be an on-chain music protocol where artists controlled governance. Instead, token voting concentrated power in early adopters and investors.

Sound: Artist-first platform with token voting. Struggled with the same plutocracy problem.

Audius: The largest music DAO (millions in treasury). Token voting led to the expected outcome: whale holders controlling proposals, low retail participation, and alignment between founders and large token holders, not between token holders and artists.

Every single one has the same pattern: token-weighted voting was inherited from general-purpose DAO templates (Snapshot, Aragon, Tally). It was never appropriate for music communities. Music is not capital-weighted. Music is contribution-weighted.

Yet no music DAO has implemented a real alternative. No music DAO uses contribution-based governance, peer evaluation, or soulbound reputation. All of them use capital-weighted voting, all of them struggle with plutocracy, and all of them have lower participation from the community members who actually create music.

Why Does This Happen? The Rational Choice

This is not stupid. It is rational.

Token voting is Sybil-resistant on a permissionless blockchain. You cannot fake tokens; you cannot split a wallet into 100 wallets and get 100x voting power (tokens have real cost). This is the only governance mechanism that works on-chain without external identity infrastructure.

Quadratic voting, quadratic funding, and other mechanisms that scale voting power with the square root of capital all fail on permissionless blockchains because they are vulnerable to Sybil attacks. If you want voting power to grow with square-root capital (n costs n^2 votes), I simply split my capital into two wallets and get 2sqrt(n/2) votes each = sqrt(2n) total votes. This beats the square-root scaling and collapses the system to linear.

The math is brutal: asymptotically, under Sybil attack, any concave voting function collapses to linear. And on a permissionless chain, the cost of a Sybil attack is just gas + the trivial cost of creating new wallets.

So token voting is the only Sybil-resistant mechanism available on-chain without external infrastructure.

The problem is that Sybil resistance is achieved by linking voting power to capital. This creates plutocracy. It is a genuine trade-off: Sybil resistance vs. plutocracy. You choose one or the other.

For the last seven years, Web3 has chosen Sybil resistance. Hence token voting. Hence plutocracy.

This is not irrational. It is the least-bad option available… if you accept the constraint that governance must be on-chain and permissionless.

But that constraint is not sacred. You can have permissioned governance (KYC, real identity) and use Sybil-resistant reputation mechanisms. You can have off-chain consensus with on-chain enforcement. You can have small groups where exit is real and Sybil attacks are impossible because people know each other.

The Governance Alternative Ancient Athens Understood: Sortition Over Voting

Two thousand five hundred years ago, Athens had the same problem.

In 500 BCE, Athens had 60,000 eligible citizens. You cannot have direct voting at that scale - it is logistically impossible. Athens needed a way to make collective decisions while preventing oligarchy (the rich and eloquent from monopolizing power).

Their solution was sortition: random selection by lot.

Citizens were randomly selected to serve on the Council of 500 (the Boule), the citizen juries (the dikasteria), and other offices. A person might serve once in their lifetime. Selection was random, not elected. The wealthy and eloquent had no advantage; anyone could be selected.

This solved several problems at once:

1. Eliminated campaign pressure - you did not need to convince 60,000 people to vote for you. You were selected at random.
2. Ensured representation - a random sample of 500 from 60,000 represents the population’s views statistically.
3. Prevented oligarchy - the rich and connected could not monopolize office.
4. Symbolized equality - everyone had equal probability of selection.

Aristotle noted: “The dikasteria contributed most to the strength of democracy.”

Athens did not use voting as its primary democratic mechanism. It used random selection and deliberation in small groups.

This is what Web3 forgot.

Modern Revival: Citizens’ Assemblies and Deliberative Democracy (2004-Present)

The 21st century revived sortition through deliberative democracy.

British Columbia Citizens’ Assembly (2004): 160 randomly selected voters deliberated on electoral reform over 11 months. The assembly recommended Single Transferable Vote (STV). A referendum in 2005 achieved 57.69% support (fell short of a 60% threshold required for implementation). The recommendation was representative and thoughtful, even if the public did not ultimately adopt it.

Ireland Citizens’ Assemblies (2016-2018): 100 randomly selected Irish citizens, stratified by age, gender, geography, and social class, deliberated on abortion, fixed-term parliaments, and climate. On the abortion question, the assembly recommended repealing Ireland’s constitutional ban. A public referendum in 2018 achieved 66.4% support - a landmark shift driven by citizen deliberation, not traditional politics.

France Citizens’ Convention on Climate (2019-2020): 150 randomly selected French citizens deliberated over seven weekends on climate policy. The output: 149 specific policy recommendations. Many were adopted into the “Projet de Loi Climat et Résilience” by the French Parliament.

In every case, the pattern is identical: small groups of randomly selected citizens, given time and expert input, produce more thoughtful, representative, and values-informed decisions than large-scale voting. The deliberation produces nuance. Voting produces yes/no.

Academic research backs this up. A 2018 Nature study (Navajas et al.) showed that aggregating just 4 consensus decisions from groups of 5 people outperformed the wisdom of 1000+ independent votes. Consensus-based small-group deliberation outperforms large-scale voting on truth-finding and decision quality.

The Core Insight: Democracy Is Not Voting

This is where the first principles matter.

Democracy is not voting. Democracy is the ability to exit and the ability to hold power accountable.

Daniel Larimer argues this in “More Equal Animals” (2021): democracy requires both the right to leave (it must be legal to exit) and the ability to leave (the cost must be practical). In a voting system where I have negligible power, I have no real exit option. I cannot exit and start an alternative community; the winning coalition will dominate it anyway because they already have concentration.

The way to preserve exit and accountability at scale is to nest it fractally - small groups where exit is cheap, federation upward where delegates remain accountable to the groups below, and the ability to fork at any level if the representatives fail to represent.

Voting, by contrast, is just a counting mechanism. It does not ensure consent. It does not ensure accountability. It just tallies preferences and implements the majority preference, even if the minority has legitimate grievances.

When voting is weighted by capital, voting becomes even more problematic. The majority becomes whatever the money votes for. This is not democracy; it is plutocracy.

Why This Matters for Music Communities Specifically

Music is the domain where earned governance matters most.

A musician’s contribution is not measured in capital. It is measured in artistic impact, innovation, collaboration, and community building. A person who released a song that defined the era is more valuable to the community than a person who bought 1000 tokens at launch.

Yet token voting inverts this. It says: whoever accumulated capital is whoever decides.

This is not a governing principle for music. This is a betrayal of it.

A music community that uses token voting is saying: we value capital over contribution. We value early adoption over artistic development. We value financial success over creative success.

This is why every music DAO has struggled. They inherited the default DAO template (token voting) without asking whether it is appropriate for music. It is not.

ZAO Fractal exists because we asked a different question: what if we built governance for musicians, by musicians, where every voting decision asks “who advanced the ZAO vision of music, art, and technology this week?”

The answer is: you get a different culture. You get 60-80% participation instead of 5-10%. You get deliberation instead of apathy. You get aligned incentives where the people earning governance power are the people doing the work.

The Historical Precedent: 500 BCE Beats 2024

This is not a new idea. Athens figured it out in 500 BCE.

We forgot. We tried voting. We tried to scale voting to millions. We tried to weight it with capital to prevent Sybil attacks.

Each iteration made the problem worse.

The solution is not a new invention. It is a return to first principles: small groups, deliberation, consensus, random selection, and nested fractal scaling. This is what worked in Athens. This is what works in modern citizens’ assemblies. This is what ZAO Fractal proves works for music.

The whitepaper documents how.

Citation Sources

• 01-foundations-deep.md (Compound 8 delegates / 50%+, Uniswap 11 delegates, participation rates 3-10%, Nakamoto coefficient)
• 04-comparative-dao-governance.md (Token-weighted voting plutocracy, Sybil resistance trade-off, quadratic voting attacks, Friends With Benefits / SongCamp / Catalog / Sound / Audius case studies)
• 01-theory-foundations.md (Ancient Athens sortition, Boule and dikasteria, Aristotle quote, British Columbia Citizens’ Assembly, Ireland Citizens’ Assemblies, France Climate Convention, Navajas et al. 2018 Nature study, Larimer’s exit thesis)

Word count: 2,148