The governance of Web3 networks and their decentralized applications has two parts: social governance and algorithmic governance.

Social Governance

Even though many processes are automated, decisions about what the rules should be or how to change them are made through public discussion and collective action by everyone involved in the network.

Algorithmic Governance

This involves rules written in computer code, like blockchain protocols or smart contracts. These rules are automatically enforced by a network of computers.

In Web3, token rewards play a key role in coordinating the network's economic activities. While we can automate some organizational tasks and enforce rules with code, the content and updates of the code depend on the input and agreement of the network's participants.

Decentralized Autonomous Organizations (DAOs) are organizations managed in whole or in part by decentralized computer program, with voting and finances handled through a blockchain. They are co-steered by the human agents acting as node operators who all have different preferences and goals. They have collective influence over the general network behaviour. It is assumed that each stakeholder in the network has their own individual self-interest, and that these interests are not always fully aligned.

Stakeholders in the network propose or vote for policy changes that will be formalized as protocol upgrades, reflecting their own self-interest. The human agents are part of the system and actively participate in the systems, either by using the services of a DAO (users), by contributing code to the network constitution (developers), or by contributing to maintain network services. In the case of the Bitcoin network, miners individually contribute to collective maintenance of a P2P payment network. In the case of MakerDAO contributors are rewarded for the collective maintenance of the Stable Token DAI. In the case of Aragon network actors are/were rewarded for the collective maintenance for a DAO platform. As a result, there are feedback-loops between the individual actors and the whole network. Since individual actions affect the system, who all have interdependencies with external events, the system as a whole evolves over time.