The history, present, and future of governance

Original Title: “Go Fork Yourself”

Compiled by: Guo Qianwen, Chain Catcher

In gaming, "speedrunning" means completing a game as quickly as possible. For example, the world record holder for Super Mario 64 is the undefeated Cheese (real name Allan Alvarez, a speedrunner), who completed the game in 1 hour, 37 minutes, and 50 seconds while singing and interacting with commentators on the Twitch streaming platform.

The cryptocurrency world has borrowed this terminology. It is speedrunning through the history of financial markets and governance.

Speedrunning is a clever but imperfect analogy. First, top speedrunners have played the game thousands of times, almost perfectly completing it each time. Second, the endpoint of a speedrun is always the same, and the speed of the game remains stable and unchanged.

But the crypto world is quite different. No one can avoid making mistakes. Many of the things done by DeFi protocols and (CeFi entities) have occurred in financial markets, making the same mistakes and learning the same lessons.

DAOs are also experiencing the same governance models—ranging from direct democracy to representative democracy, from direct shareholder voting to board management—these methods have been applied in local and national governments and companies. They simply do it faster, compressing thousands of years of experimentation into less than a decade.

Andy Hall and Porter Smith call it "light-speed democracy." This name is closer to the mark.

However, its development should not stop there. Online advertising initially replicated print advertising in the form of banner ads, but then evolved into a richer and more sophisticated toolkit that could not be realized offline. Similarly, DAO governance should not be limited to offline models.

DAO governance is more like an ecological process: governance is native to the network and continuously evolves.

Our goal should not be to replicate offline governance online—after continuous trial and error, network-native organizations cannot and should not operate like geographically-based governments, as the limitations they face are different. Once the online governance model evolves to a certain extent, it should be different and superior to offline models, as it has advantages in speed, scale, granularity, programmability, composability, and its binding to the internet and blockchain.

What if we reverse the model?

What if the goal of DAO governance is not to reach a consensus on the limited choices communicated by the management but to encourage people to generate disagreements in organizational decisions? If large communities become unwieldy and difficult to collaborate, they should continue to divide into smaller, more effective communities to co-create value.

The future of DAO governance may be forking: through governance, encouraging disagreements among people, creating sub-communities—where members share common goals and can create projects that align with their vision. In this sense, forking is the ultimate form of decentralization. Thus, governance becomes the cornerstone of the social landscape, where people can find companions and share and realize common interests.

In its simplest form, this governance is a decentralized management that incentivizes communities to share and rate preferences. Healthy disagreements form sub-DAOs, encouraging the ecosystem to evolve.

From this perspective, governance mechanisms are no longer so defensive—spending less energy on how to protect the entire DAO from attacks—they become more aggressive and interesting—focusing more on how to make governance a means of entertainment, social discovery, and dissemination.

Of course, there is no one-size-fits-all model for DAO governance. We do not advocate that every DAO rushes to fork immediately. In fact, forking should be the last resort. What we want to say is that different types of governance can allow different organizations to have different standards, applicable to what they consider important for their systems. Essentially, biodiversity can protect the entire ecosystem.

Of course, this model also faces challenges. The most obvious ones are reduced liquidity and increased complexity. We will discuss this later.

If experimentation and trial and error are the necessary paths, then we should also gain insights. In a favorable situation with fewer constraints, we should experiment in different directions at the endpoint, using new tools to create new internet-native opportunities. Perhaps one day, these new models will, in turn, influence the way humans coordinate offline. Perhaps the boundaries between online and offline will blur to the extent that governments and companies will adopt new models born from this evolution.

Today, we will explore the history, present, and suggested future of governance.

• A Brief History of Governance
• The History and Challenges of DAO Governance
• Forking Governance
• How Governance Works
• Governance as a Social Activity
• The Evolution of Internet-Native Governance

A Brief History of Governance

Governance is the process of overseeing the management and direction of affairs. In this process, a group of people decides what the group should do, ensuring that decisions can be made.

Humans have been contemplating governance since the dawn of time. In the book Pieces of the Action, Vannevar Bush writes:

"When Adam and Eve formed a partnership, that was the first organization in history. But the key governance relationships and rules have not been fully realized to this day. After Adam, humanity began to construct increasingly complex organizations to carry out affairs."

Historically, humans have been ruled by aristocracy, where power was concentrated in one person: a king, queen, empress, emperor, or dictator. Whatever decrees they issued, the people had to obey. Oligarchy has also been quite popular throughout history, with the main difference being that power is concentrated in a small group of people. Dictatorship was also prevalent in early history because it is the simplest model of governance—one person can make all decisions—but history shows that this system is flawed.

In the 5th century BC, the ancient Greeks implemented a new form of governance (tribal societies had practiced it earlier): democracy. In ancient Athens, one of the earliest homes of democracy, all citizens—i.e., native-born free men—were required to actively participate in government work. Those who did not participate could be fined and sometimes even splashed with insulting red paint.

The way Athenian democracy worked was similar to jury duty. Each year, 500 citizens were selected from all citizens of ancient Athens, and they had to actively serve in government for one year. During this year, they were responsible for creating new laws and managing all parts of the political process. When new laws were proposed, all citizens of Athens had the opportunity to vote on them. (To vote, citizens had to attend the assembly on the voting day.)

This form of governance, where every citizen could vote on every decision, is called direct democracy. Each citizen had their own voice, and everyone had to follow the majority's decision. Direct democracy worked effectively in ancient Athens because of its small scale and local nature, but even so, Athenians eliminated 70% of eligible voters (women, slaves, and immigrants) to maintain small groups.

Although the ancient Greeks introduced democracy 2,500 years ago, and Rome was governed by a republican government from 500 BC to 27 AD, autocracy remained the dominant form of governance globally until the enactment of the U.S. Constitution in 1789, with few exceptions.

While direct democracy can make a city run smoothly, the United States needed a governance model that could handle the 13 colonies (eventually 50 states) spread across a vast area.

As James Madison wrote in Federalist No. 14: "In a democracy, the people can see the government in person, executing affairs; in a republic, the people manage the government through representatives. Thus, democracy is limited to one place. Perhaps a republic can be extended over large areas."

New environments require new forms of governance. The Founding Fathers of the United States established a republic, a form of representative democracy. The people elect their representatives, who propose and pass legislation.

Since then, the world has inevitably embarked on the rocky road of representative democracy.

A 2019 Pew Research report found that 57% of countries with populations over 500,000 are democracies. It is evident that modern democracy is representative democracy. The most common form of governance in companies is also representative democracy. Shareholders elect a board of directors, the board hires and fires the management team, and the management team hires and fires employees, managing the day-to-day operations of the company.

Although representative democracy also has flaws, it seems not as bad as previous forms of governance. As British Prime Minister Winston Churchill famously said, "Democracy is the worst form of government—except for all the others that have been tried."

The history of governance is a series of attempts to find better solutions than those of predecessors, applicable to the context and capabilities of the current era. New tools, challenges, and constraints open the door to new models. The ultimate form of governance will certainly not stop at existing models.

Governance can and should continuously evolve. States and companies find it difficult to experiment with new governance forms, but DAOs benefit from the internet and can quickly experiment.

The History and Challenges of DAO Governance

Unfortunately, so far, the governance models of DAOs are almost still mimetic versions of state and corporate models.

Today, most (if not all) cryptocurrency governance mechanisms are implemented around some concept of "one token, one vote," similar to token-weighted direct democracy or shareholder democracy.

DAOs can also choose not to adopt a purely one-token-one-vote system—mechanism designers use novel on-chain features to enhance old models without trying new ones. Some popular enhancements include:

One Wallet, One Vote: Similar to direct democracy, this voting method aims to give each member equal voting rights. However, it faces challenges, such as voters with insufficient information or low participation having the same voice as those who are more informed and enthusiastic, and one person can easily set up multiple wallets to vote.

Quadratic Voting: Aims to reduce the voting power of large shareholders, but they still have more say compared to small shareholders. This quadratic voting system increases the cost of each vote; one vote may cost one token, but two votes cost four, three votes cost eight, and so on.

Proof of Participation: Proposed by people including Vitalik Buterin, proof of participation only grants votes to those who actively participate in the DAO or protocol, making it more difficult and expensive to launch governance attacks.

There are other methods, but most primarily focus on deploying protective layers to enhance productivity and mitigate or delay the impact of governance attacks—so-called optimistic governance. All of these are important improvements, but they do not answer a key question: Does the underlying governance framework adopted incentivize positive behavior when dealing with complex tasks?

Our guess is that for the vast majority, if not all, DAO governance models, the answer is no.

Initially, most on-chain governance systems were designed to manage very simple decisions: for example, whitelisting collateral tokens, modifying parameters, activating or deactivating oracles. The tasks at hand for DAOs were clear, and contributors only needed to maintain normal operations or make slight improvements.

However, human ambition is boundless. With significant economic success, protocols began to expand the scope of governance, and simple matters gradually became complex, while direct democracy struggled to address complex issues effectively.

Take MakerDAO as an example. It initially intended to be a simple permissionless vault for storing cryptocurrency collateral and obtaining $DAI financing, but it quickly attempted to transform itself into a lender focused on real-world sustainability.

Using vast sums to combat climate change may be the most complex test for any human organization. Unsurprisingly, Maker attempted this but performed poorly. Questionable institutions continuously boasted about their "green" initiatives to secure massive funding. They heavily promoted their initiatives across governance channels for months, legitimizing their endeavors in the eyes of token holders. In a flat democratic system, without a clear system of checks and balances and specialization of tasks, it is difficult to distinguish between good and bad actors. Ultimately, Maker decided to purchase U.S. Treasury bonds—clearly neither its original intention nor an environmentally friendly investment strategy.

Getting numerous DAO members to evaluate complex proposals with a professional and fair attitude regarding their economic impact is often impossible.

These systems are still winner-takes-all. Members with enough voting power who can bear the management risks of voting can decide the direction of the entire community without incurring other costs. Similarly, transaction fees are a significant issue for many small token holders; unfortunately, in the crypto world, most token holders are more concerned with yield farming than voting matters. We guess that as long as one can secure 10-20% of all votes, they can determine the fate of most DAOs; thus, the voice of the winner may not represent all members.

Low participation is the most dangerous situation in modern democracy. DAOs, composed of DeFi protocols, can allocate vast financial resources, exacerbating this issue.

Due to the complexity of the problems, voter apathy, and overly simplified governance frameworks, it is not difficult for a small group of voters to dominate most protocols. Those voters holding significant assets have no reason to invest additional funds to rally the team—they only need to purchase tokens to vote and can benefit from supporting proposals while dispersing potential losses.

For instance, an unscrupulous individual has a protocol that receives $1 billion in funding from another protocol to support his dubious project. He can purchase $50 million worth of governance tokens from the latter protocol and then vote to benefit his protocol. If those $50 million tokens can enable him to take control, he could even abscond with the $1 billion. In such a one-token-one-vote system, the temptation to become a bad actor is too great.

Voting becomes a feast for lobbyists.

But what if his $50 million (5%) can at most only secure him 5% of the funds? He wouldn't bother with such actions. After all, the amount he ultimately receives remains the same.

Not just in DeFi, large amounts of capital lead to high-profile and high-risk actions. DAOs face three broader application issues: centralization, prioritization, and vulnerability.

Centralization. If a DAO is like a fanatical group bowing before a grand archbishop, is it still a DAO? People often complain that the one-token-one-vote majority governance system of DAOs is centralized, inevitably leading to power inflation for founders and whales.

But the centralization issue is also a product of the process because DAO governance completely mimics traditional shareholder governance, where key teams must obtain board approval to act, rather than encouraging the community to raise questions and actively voice concerns. Therefore, decentralized governance has not yet unlocked the most exciting aspects of Web3: fostering mutual trust among members and enabling them to collaborate creatively. Native DAO governance should, in turn, allow members to submit proposals rather than follow suggestions put forth by core members.

Prioritization. Generally, if you run an investment DAO and need to decide whether to invest in certain protocols, you would conduct a resolution vote. We do not do this; instead, we hold a competition to see which protocols are more popular.

This first means that the criteria for comparison become more granular, allowing you to compare people's varying levels of enthusiasm for protocols. Secondly, it can reduce public negativity toward vetoing protocols, as people's votes in a competition are all positive. Thirdly, the marketing costs for projects seeking approval will be higher. Ultimately, you can distribute funds to different projects based on voter support.

Vulnerability. Vulnerability is unrelated to the achieved decentralization and work efficiency. If a system can be easily attacked, it means all construction work and resources will be exposed. In the crypto world, such attacks are more frequent, so effective DAOs should develop systems to respond.

To address these three issues, the simplest way is to change the submission and voting processes. Let your community submit proposals and vote on what they like. This sounds very simple, but why have we not seen such systems implemented?

Our readers might scoff, "Letting community members compete with each other to decide the direction of the community—will only exacerbate division and disagreement, forcing experts to heed the opinions of outsiders."

Yes, indeed, because it should be so. We need forks.

Forking Governance

The answer is forking. Forking refers to creating a new version of a system under new ownership. For example, tweaking DNA can lead to species mutations, producing entirely new animals, but the origin of the species remains unchanged.

Forking is not a new concept in the crypto world.

The most famous fork could be Ethereum. After the DAO was attacked in 2016, Ethereum was forked to trace funds back to user accounts. Despite the controversy, it perfectly demonstrated how to violate the immutability of the blockchain to create a better model based on social consensus. In fact, subjectivity is the foundation of governance systems. In retrospect, this process resembles a linear growth evolution, with countless almost random forks in between, but ultimately only the most suitable ones remain. Forking is the process of increasing probabilities.

In short, forking is a network-native political function that can never be realized in traditional states or governments: if you disagree with the decisions of the existing government, you can choose to start your own online empire with its own politics and currency. In real life, this is not feasible because you have to manage land, wage wars, and build social consensus. But in the online world, you only need to copy code and find companions.

Thus, the network allows us to "speedrun" government building.

The most common criticism of democracy is that the politically uninformed public does not know what is best for them, and democracy is also constrained by cultural conflicts—groups opposing each other and the dull consensus of committees—only the most conservative and compromising proposals can pass, while those genuinely willing to address important issues are sidelined.

Forking provides a third path. You can start your own community and build your own vision. In the crypto world, the open-source, permissionless environment enables small teams to collaborate and execute work quickly. Value belongs to tokens rather than projects. Typically, new trends first attract media and entertainment industries before shifting to more complex enterprises. A prime example here is the Nouns NFT project, which showcases the potential of DAO forking.

Every day, Nouns auctions an NFT, and for the next 393 consecutive days, people buy a Noun for 80-100 ETH ($136,000-$170,000). The winner receives the Nouns NFT and becomes a member of NounsDAO, managing a treasury containing 26,350 ETH ($45.3 million). Members of NounsDAO can propose and vote to allocate funds from the treasury, which are typically used to create projects based on Nouns under the cc0 license without permission.

A notable example is Nouns Vision, a luxury physical sunglasses line based on the iconic Nouns glasses design. In February, salvinoarmati proposed creating a prototype for the sunglasses, and the DAO granted him 5 ETH. In April, the prototype was completed, and DAO members (also known as Nounders) proposed allocating 7.2% of the treasury funds to bring 500 pairs of sunglasses to market for the first 500 Noun holders. 48 members voted in favor, while 8 voted against, and the proposal passed.

Beyond these 500, the creator of the sunglasses sold them to non-NounsDAO members for 0.44 ETH each, and the sales proceeds did not return to the NounsDAO treasury.

Why would a DAO fund a project that uses their IP without paying?

In the era of TikTok and memes, companies often rely on user-generated content for marketing, rapid dissemination, and decentralized brand building. This user-generated content is essentially forking—users re-creating stories based on core IP. Nouns makes it as easy as possible for people to create using their IP and profit from it—even paying from their own treasury.

Nouns is making a very 21st-century internet bet: by incentivizing everyone to use its IP for free, more value will return to the core IP. To bring more value to the brand, it relinquishes control over projects like Nouns Vision and short-term revenue, ultimately leading to higher auction prices and greater demand for projects developed by the DAO, which is the additional value it can reap. In the crypto world, forking is often seen as an alternative version of the blockchain, but with new tokens. However, Nouns demonstrates another capability of forking: a positive fork that accumulates value back to the original token and brand.

Nouns is not an isolated case.

Ethereum Rollups (like Optimism and Arbitrum) can also be considered a form of forking, enhancing Ethereum's core functionality—execution—creating more value for its tokens.

At the same time, EigenLayer allows ETH stakers to stake again—collateralizing their tokens to provide validation services beyond ETH transactions for rewards. Essentially, if a project splits into multiple projects using the same token, even if the projects hate each other and disagree, they are still creating value for that token.

Finally, let us ponder a question.

Should optimal governance represent everyone's views? Or should it rely on the public to build consensus among parties to pursue common goals? Under the current system, broad consensus is not feasible; the optimal solution is merely a consensus reached by small power groups that influence the entire community.

But when forking becomes an option, we have more choices online.

First, internet-native governance allows us to choose to participate in or leave the system, encouraging differing opinions and enabling parties with differing views to find each other and pursue common goals.

Second, forking also means DAOs can continuously split into sub-DAOs, which are smaller and easier to coordinate compared to large systems—more like the direct democracy of 500 Greeks rather than the representative democracy of millions of Americans.

Third, forking allows larger groups to continually split into smaller groups, similar to mitosis—some of which are destined to grow independently—this governance now means individuals can build relationships, communicate with others, and listen to others' ideas to determine common goals. These actions can yield little effect in large democratic nations.

In this context, governance is no longer an activity that occurs every few years at the polls but rather a continuous social activity.

Governance as a Social Activity

In other words, internet-native governance not only reflects decisions but is also the decision-making process itself, in which people actively build relationships and feel excited about common goals, thus achieving those goals more effectively. Therefore, the key is not voting but the discussions that build consensus. This is the characteristic of DAOs.

For example, Twitter or any other mainstream Web2 social network serves as a governance platform. Every like is a vote for the content, indicating your agreement with it, your desire to see more similar content, and your hope that others with similar interests can also see it. Most importantly, the network institutionalizes the two main components of internet-native governance: forking (through quoting, retweeting, copying, pasting, and TikTok re-creation) and forums (incentivizing participants to raise objections or build consensus in comments).

Thus, successful governance can be understood as a fun large online game based on team building, with competition and reward mechanisms. The gamification of governance means that the focus of governance is no longer on outcomes but on the governance process itself—just as playing a game emphasizes the gameplay.

DAO governance can draw lessons from these social platforms while unleashing the potential of Web3. Good governance is, in fact, a Web3 social network—where people generate disagreements and discussions based on social graphs, find partners with shared interests, and pursue common goals within the community.

How Forking Works

A quick answer is that dissenters should be able to find each other on-chain and communicate; these surface relationships are more important than the decisions themselves.

There are two types of forking: governance forking and protocol forking.

Governance Forking.

Suppose we have a DAO X, and you can vote using its token $DAO X. In a simplified majority governance system of one token, one vote, the winning group can decide which policies to implement. The final passing ratio does not matter (whether it is 50.1% vs. 49.9% or another value).

But in a world that allows forking, the group that loses with 49.9% would receive 49.9% of the assets from the DAO treasury, which would be converted into a newly minted sub-governance token $DAO-X-LOSER to decide matters. Similarly, the remaining 50.1% of the assets would undergo the same process.

Forking would allow the two groups to continue their respective governance processes, allocating financial resources based on the community's passing ratio. Forking enables experimentation and evolution to occur while reducing the existing risks of the protocol. Clearly, the forking process can continue indefinitely.

This type of forking system will face obvious challenges. Hackers will not attack the entire treasury; after all, it is difficult to breach, but will instead launch smaller governance attacks, nibbling away at a small portion of the treasury, which, while not much, is still significant.

One important thing is: people can only vote with the tokens they control, and the final result may be that they abandon their existing tokens and leave together or fork those tokens to do things that pose the least risk to the core DAO but may yield benefits.

Using the previously described Maker case as an example. If 10% of the DAO's members are climate activists or climate experts who are very enthusiastic about climate-positive loans, they could fork 10% of the treasury assets to lend to climate-positive borrowers and focus all their energy on building the system to guarantee such projects. In addition to the current 10% funding, they might also attract external funding that values the Maker infrastructure investment, willing to allocate funds specifically for climate-positive loans rather than other Maker loans. They could establish a token exchange or fee system, allowing the sub-DAO to rely on Maker, and Maker could profit from potentially high-risk loans.

Even without malicious actors, forking could excessively fragment the liquidity of the protocol, which would have unexpected negative impacts on each participant. Who would want to pick tokens for swapping among 100 different Uniswap sub-types?

Solutions should be situational. In some cases, the DAO might limit the maximum proportion of the treasury used for forking.
It could also require sub-DAOs to pay fees to the DAO, allowing forking to gain product control but preventing them from absconding with funds. The costs of forking should be minimized as much as possible but should not disappear entirely.

Our consistent principle is that potential benefits should outweigh potential harms. Malicious actions should incur high costs, but the forking process should be easy.

From the perspective of investors or participants, such a system can more clearly demonstrate how internal organizations within the DAO manage resources, providing more specific choices for their investments. Investors or participants can always access the value accumulated after all forks and can choose to engage only with a specific sub-group, thereby increasing internal unity.

However, most of the time, governance forking should not be the choice. DAO members should not only vote on proposals but should also propose, adjust, submit, and readjust proposals before they enter binding votes.

Protocol Forking

For example, if two-thirds of a small group reject a proposal for various reasons. One-third oppose the budget, and another third oppose the timeline. However, what if different factions adjust the details, submit different versions of the proposal, vote on their favorite proposals, and then ultimately vote for and against that proposal on-chain? Two things would happen.

First, the community would have a clearer understanding of the ideal implementation of a proposal. Currently, Lido DAO is resubmitting various versions of treasury diversification proposals to see which can withstand the test of time; if the community is allowed to submit suggestions, it can effectively find the best proposal.

Second, if a more popular version of the proposal emerges, the 67% opposition might drop to 20% or 30%. This "proposal forking" process can bring about stronger social consensus, as parties can actively negotiate to achieve common goals, revealing the commonalities of the community rather than points of disagreement. Proposals can pass, and the community may thus be happier.

All of this can be achieved on governance platforms like jokedao, where incorporating community feedback into the governance process can yield such effects. When people can express their insights, they will strive to provide the best answers to win.

In a sense, building applications on certain blockchains is just such a scene.

We are already in a world of continuous governance forking—we just need to fully accept this reality. Continuous forking mechanisms can maximize the professionalism of work, enhance the possibility of economic independence, and gradually expand behavioral differences between sub-groups, thereby maximizing group belonging while maintaining a common vision.

The Evolution of Speedrunning

DAOs do not need to pursue perfection; they only need to strive for improvement.

Recent debates about cryptocurrency use cases have exposed an ironclad fact: Web3 is not as adept at doing the business of Web2 companies as Web2 companies are.

But one thing is certain: real-world companies mostly cannot fork, cannot easily exchange equity, and cannot treat governance as entertainment and social activity.

But DAOs can. Because they exist on the internet, unbound by geographical or physical constraints, they can evolve faster, experiment more, and fork into countless competitive and cooperative sub-groups, developing at a speed greater than that of humans or existing institutions. DAOs do not need to pursue perfection; they should focus on increasing "mutation," allowing the market to favor the fittest, and even mix and match "mutations" in digital reproduction.

Letting DAOs operate like internet-native organizations, free from constraints, can accelerate the above processes on two levels.

System Level. More experimentation with governance models themselves can help better models emerge more quickly.

Sub-DAO Level. More forking, backed by the resources of the parent DAO and supported by participants in sub-DAOs with aligned goals, will lead to the flourishing of small, effective (and interesting) actions.

DAO governance models that encourage dissent and forking may directly or indirectly increase value. They may provide an intuitive defensive mechanism to prevent bad actors from exploiting existing models to seize large portions of the pie. Relinquishing control and allowing the system to develop naturally will create fertile ground for various unpredictable adaptive changes.

Perhaps the most significant impact is not economic but the increase in individual choices. In Politics, Aristotle said, "It is evident that this government is the best in the world because, regardless of a person's identity, he/she can take the best actions and live happily."

In the real world, no one can always take the best actions and live worry-free. Life involves compromises and trade-offs. But in the online world, within rich, niche, closely collaborating, decentralized sub-communities, people can play, work, connect, invest, and realize their visions and happiness.

Of course, new models still need exploration, seeking ways to operate and filling in gaps. We hope this article can spark some discussions. Of course, discussion is one thing, and practice is another.