Consensus algorithms are the cores of blockchains. It is a rule that all nodes follow to maintain a consistent and ever-growing ledger. This rule is deliberately written in code so that the nodes could follow it automatically and unambiguously. Any violation of this rule will be considered as an adversarial behavior and could receive economical punishment. In fact, this is exactly where the philosophy of DAO roots in. Ideally, the operation of a DAO should also be deliberately written in codes and ran automatically, which as we know is currently impractical. However, rules (let's just use the term "consensus") are still crucial to a DAO. The consensus should be much stronger than a vision, a goal, or a manifesto. It is a set of practical rules that all participants should follow, including

1. The governance mechanism, which defines their rights in the DAO and how to exercise them;

2. Workflows and dismissing rules, which describe the obligations of the members with specifications of how to fulfill them;

3. Incentive mechanism and finance, which specify the reward of the contributions and the possible bonus in the future. The consensus should be as specific and unambiguous as possible. Although it is not yet possible to write everything down as smart contracts and automatically run them on blockchain, it is desirable to use the blockchain technology, or any suitable technology, to remove as much human influence as possible. With such consensus, it is also made clear how to onboard or contribute to a DAO. All DAOers implicitly sign a contract to follow the consensus when joining the DAO. They should follow the workflow to get rewards and fulfill their obligations to earn rights to govern. At this moment, very few DAOs have consensuses that meet this standard. However, I believe such consensuses are musts for all DAOs, which lead to a conclusion: in fact, all DAOs, either eventually derive a consensus that meet this standard, or would fail in a long-term.

Fork is arguably a great invention in blockchain technology. Although sometimes considered as a bad event in many public blockchains, it might be in fact the best outcome for an already broken consensus. A fork usually happens when a group of people who once shared the same vision, now have strong disagreement and cannot work together. The key advantage of forks in blockchains is that both forks will inherit all on-chain legacies. For example, in Bitcoin, the legacies include users, ledger, and protocols, which are inherited by both forks. Both forks could strive and the market will decide the value of them. For instance, the total market cap of Bitcoin increased right after the fork of Bitcoin (BTC) and Bitcoin Cash (BCH) due to the fact that the market believes that Bitcoin could achieve more if they go separate ways. Indeed, the off-chain legacies like community, public recognition, collaborators, etc., cannot be fully inherited. However, this is already a huge improvement over traditional organizations where the splitting could be much harder and uglier. In my opinion, forks of DAOs should also be made possible and even encouraged. This conclusion directly follows my other thoughts: 1) DAOs can have different forms; 2) DAOs should have clear and compulsory consensuses ; 3) DAOs will gradually shape their consensuses by governance if they do not have them now. Then, in the process, DAOers will certainly have disagreements with the consensuses. For example, as a DAO thrives, some DAOers would prefer to work in a more organized fashion while the others might prefer a more relaxed form. In traditional organizations, it is natural to downplay the conflict and try to hash out a solution that fits the interest of the majority. Then, the rest should subordinate to the majority. However, in DAOs, it makes no sense to force anyone to follow a rule that he/she no longer believes in and contributes to someone else's goal. In case of disagreement, it is better that the DAO forks into two DAOs which have different consensuses but remain bonded, than to have one DAO with internal conflicts that grow harsher over time. Even if divergences are mitigated, the conflict of interests remains, which could still lead to the loss of motivation and eventually the failure of the DAO. Autonomy is the key success factor of DAOs. All DAOers should either fully believe in the consensus of a DAO and dedicate to it, or leave it and create or find another DAO that they would like to believe in.  Moreover, the relationship of DAOs should be reconsidered. An individual could simultaneously participate in multiple DAOs and it is possible to have multiple DAOs with the same members but different consensuses. Essentially, DAOs are not rivals to each other but comrades: comparing to other forms of organizations, it is easier for DAOs to collaborate since they share the same organizational form. Then, among all DAOs, a fork would be the closest buddy that a DAO could have, since it once shared the same consensus. As a matter of fact, the forks of a DAO could collaborate in a form not possible in other organizations: a member could simultaneously work in both DAOs while the contribution could be recognized by both DAOs and rewarded with both tokens with different weights.  In general, forking should be made a viable solution for all DAOs and the process could be made automatic and painless. I might even consider the freedom of forking as a key of success of DAOs in the future.

The ultimate form of DAOs has been widely discussed. It is mostly agreed that DAOs should not be very hierarchical. Then, some DAOs follow the Bankless DAO's track to create guilds or silos, in which the members are categorized by their skills. A project proposal will then reach out to the guilds and recruit talents to complete their tasks. Others believe that DAOs consist of smaller units called "Pods",smaller groups of people that share similar interests or skills and autonomously formed within a DAO. My thoughts differ from both of them. I believe that there should be no other forms of sub-organizations within a DAO. DAO is merely a group of people sharing the same consensus and act accordingly to it. Hence, in many cases, the guilds/silos/pods are also DAOs as long as they follow a consensus that others do not follow. In that sense, a DAO could consist of many smaller DAOs, which seems to form a hierarchical structure in DAOs. However, the relationship between DAOs could be much more complicated thanks to the distinct features of DAOs. Two DAOs could share members, share visions, collaborate in tasks, hold each other's tokens, participate in each other's governance, or even share all information properties if they are forked from the same root. In other words, the members of a larger DAO X could simultaneously participate in many smaller DAOs, including DAO Y, which is funded and partially governed by X. However, Y is not necessarily a sub-DAO of X or has any organization pledge to X. Then, the only thing that differs two DAOs are their consensuses, which consist of governance, finance, membership management, and workflows. These components can be broken into smaller pieces, namely protocols, which do not come from anywhere. A reasonable guess would be that some protocols are popular and consensuses are simply the combinations of various protocols. Then, the "sub-DAO" Y is merely X with some additional protocols. In fact, the kind of "sub-DAO" is also a protocol, which could be universally used to regulate how such sub-DAOs could be created, funded, and governed. All relationships between DAOs could also be abstracted into protocols, e.g., forks, sub-DAOs, collaborations, retailers, etc. The protocols will be the edges that connect the DAOs in a DAO universe.

Consensus algorithms are the cores of blockchains. It is a rule that all nodes follow to maintain a consistent and ever-growing ledger. This rule is deliberately written in code so that the nodes could follow it automatically and unambiguously. Any violation of this rule will be considered as an adversarial behavior and could receive economical punishment. In fact, this is exactly where the philosophy of DAO roots in. Ideally, the operation of a DAO should also be deliberately written in codes and ran automatically, which as we know is currently impractical. However, rules (let's just use the term "consensus") are still crucial to a DAO. The consensus should be much stronger than a vision, a goal, or a manifesto. It is a set of practical rules that all participants should follow, including

1. The governance mechanism, which defines their rights in the DAO and how to exercise them;

2. Workflows and dismissing rules, which describe the obligations of the members with specifications of how to fulfill them;

3. Incentive mechanism and finance, which specify the reward of the contributions and the possible bonus in the future. The consensus should be as specific and unambiguous as possible. Although it is not yet possible to write everything down as smart contracts and automatically run them on blockchain, it is desirable to use the blockchain technology, or any suitable technology, to remove as much human influence as possible. With such consensus, it is also made clear how to onboard or contribute to a DAO. All DAOers implicitly sign a contract to follow the consensus when joining the DAO. They should follow the workflow to get rewards and fulfill their obligations to earn rights to govern. At this moment, very few DAOs have consensuses that meet this standard. However, I believe such consensuses are musts for all DAOs, which lead to a conclusion: in fact, all DAOs, either eventually derive a consensus that meet this standard, or would fail in a long-term.

Fork is arguably a great invention in blockchain technology. Although sometimes considered as a bad event in many public blockchains, it might be in fact the best outcome for an already broken consensus. A fork usually happens when a group of people who once shared the same vision, now have strong disagreement and cannot work together. The key advantage of forks in blockchains is that both forks will inherit all on-chain legacies. For example, in Bitcoin, the legacies include users, ledger, and protocols, which are inherited by both forks. Both forks could strive and the market will decide the value of them. For instance, the total market cap of Bitcoin increased right after the fork of Bitcoin (BTC) and Bitcoin Cash (BCH) due to the fact that the market believes that Bitcoin could achieve more if they go separate ways. Indeed, the off-chain legacies like community, public recognition, collaborators, etc., cannot be fully inherited. However, this is already a huge improvement over traditional organizations where the splitting could be much harder and uglier. In my opinion, forks of DAOs should also be made possible and even encouraged. This conclusion directly follows my other thoughts: 1) DAOs can have different forms; 2) DAOs should have clear and compulsory consensuses ; 3) DAOs will gradually shape their consensuses by governance if they do not have them now. Then, in the process, DAOers will certainly have disagreements with the consensuses. For example, as a DAO thrives, some DAOers would prefer to work in a more organized fashion while the others might prefer a more relaxed form. In traditional organizations, it is natural to downplay the conflict and try to hash out a solution that fits the interest of the majority. Then, the rest should subordinate to the majority. However, in DAOs, it makes no sense to force anyone to follow a rule that he/she no longer believes in and contributes to someone else's goal. In case of disagreement, it is better that the DAO forks into two DAOs which have different consensuses but remain bonded, than to have one DAO with internal conflicts that grow harsher over time. Even if divergences are mitigated, the conflict of interests remains, which could still lead to the loss of motivation and eventually the failure of the DAO. Autonomy is the key success factor of DAOs. All DAOers should either fully believe in the consensus of a DAO and dedicate to it, or leave it and create or find another DAO that they would like to believe in.  Moreover, the relationship of DAOs should be reconsidered. An individual could simultaneously participate in multiple DAOs and it is possible to have multiple DAOs with the same members but different consensuses. Essentially, DAOs are not rivals to each other but comrades: comparing to other forms of organizations, it is easier for DAOs to collaborate since they share the same organizational form. Then, among all DAOs, a fork would be the closest buddy that a DAO could have, since it once shared the same consensus. As a matter of fact, the forks of a DAO could collaborate in a form not possible in other organizations: a member could simultaneously work in both DAOs while the contribution could be recognized by both DAOs and rewarded with both tokens with different weights.  In general, forking should be made a viable solution for all DAOs and the process could be made automatic and painless. I might even consider the freedom of forking as a key of success of DAOs in the future.

The ultimate form of DAOs has been widely discussed. It is mostly agreed that DAOs should not be very hierarchical. Then, some DAOs follow the Bankless DAO's track to create guilds or silos, in which the members are categorized by their skills. A project proposal will then reach out to the guilds and recruit talents to complete their tasks. Others believe that DAOs consist of smaller units called "Pods",smaller groups of people that share similar interests or skills and autonomously formed within a DAO. My thoughts differ from both of them. I believe that there should be no other forms of sub-organizations within a DAO. DAO is merely a group of people sharing the same consensus and act accordingly to it. Hence, in many cases, the guilds/silos/pods are also DAOs as long as they follow a consensus that others do not follow. In that sense, a DAO could consist of many smaller DAOs, which seems to form a hierarchical structure in DAOs. However, the relationship between DAOs could be much more complicated thanks to the distinct features of DAOs. Two DAOs could share members, share visions, collaborate in tasks, hold each other's tokens, participate in each other's governance, or even share all information properties if they are forked from the same root. In other words, the members of a larger DAO X could simultaneously participate in many smaller DAOs, including DAO Y, which is funded and partially governed by X. However, Y is not necessarily a sub-DAO of X or has any organization pledge to X. Then, the only thing that differs two DAOs are their consensuses, which consist of governance, finance, membership management, and workflows. These components can be broken into smaller pieces, namely protocols, which do not come from anywhere. A reasonable guess would be that some protocols are popular and consensuses are simply the combinations of various protocols. Then, the "sub-DAO" Y is merely X with some additional protocols. In fact, the kind of "sub-DAO" is also a protocol, which could be universally used to regulate how such sub-DAOs could be created, funded, and governed. All relationships between DAOs could also be abstracted into protocols, e.g., forks, sub-DAOs, collaborations, retailers, etc. The protocols will be the edges that connect the DAOs in a DAO universe.