Original Title: 6 decentralized governance trends for 2025
Original Author: Andrew Hall, A16Z
Original Source: https://a16zcrypto.com/posts/article/6-decentralized-governance-trends-for-2025/
Translated by: Tom, Mars Finance
Six Decentralized Governance Trends: Looking Ahead to 2025
As the position of the online world in our lives becomes increasingly important, the critical collective decisions we need to make online will also increase. For this reason, 2025 is worth looking forward to in the field of decentralized governance: DAOs will continue to explore new ways to enable anonymous token holders to govern together; meanwhile, large institutions like Blackrock, State Street, and Vanguard are also trying to encourage customer participation in online shareholder voting; AI companies like Anthropic, Meta, and OpenAI are using citizen assemblies to set usage rules for large language models (LLMs).
In this strong momentum, here are some themes I will continue to watch and participate in next year.
1. Websites that facilitate delegated voting
Theme Background: Delegated voting is crucial for online governance, but how to enable voters to choose their delegates wisely remains a challenge. Research indicates that user-friendly delegated voting sites can increase participation rates by over 20%, but when there are many delegates to choose from (which is a good phenomenon), how to present this information without further concentrating power becomes crucial. Currently, we are uncertain about the best way to present information and who should be responsible for such important displays. The year 2025 will be an important year for exploring how to present delegated information and track its impact.
2. AI helps complete voting delegation
Theme Background: Can AI agents help users find delegates that align with their views? Perhaps users can converse with AI agents, answering questions about their personal values and goals; the AI agents would then review the voting records, declarations, proposal content, and forum discussions of the delegates to determine which delegate is most suitable. With the acceleration of generative AI competition, 2025 will be an excellent experimental period for the intersection of AI and governance, which is exciting.
3. AI itself acts as a delegate
Theme Background: A more radical idea is whether we can develop AI agents that understand user preferences and can vote on behalf of users. Can they discuss in forums and even co-write proposals with other AI agents? True AI delegates may still be a distant reality, but next year is a good time to explore this model.
4. Smarter participation incentives
Theme Background: Experience and research show that simply issuing one-time airdrop rewards is often insufficient to guide people to participate deeply and long-term in decentralized projects. In 2025, we will see more interesting explorations in 'beyond airdrops.' If the project itself needs to incentivize activities with clearly defined economic value (such as trading), direct subsidies can be used; if the project wants to encourage harder-to-measure behaviors, longer-term incentive mechanisms can be considered, such as repeated rewards combined with time locks.
5. Better funding models for public goods
Theme Background: Many projects require the assistance of community members to build ecosystems. However, existing grant and retroactive rewards methods often fail to effectively encourage high-risk, high-reward ideas: members tend to prefer projects with a higher chance of success that can secure funding, rather than those they truly believe in. A reasonable approach is to explore a venture capital-like model for public goods, providing sufficient upfront funding to projects, and offering higher returns once the community recognizes the project's success. The year 2025 may become the 'take-off year' for such innovative funding schemes.
6. More attempts at sortition
Theme Background: AI companies like Anthropic, Meta, and OpenAI have explored using random sampling to determine AI usage rules. Random sampling refers to randomly selecting a portion of users to participate in structured discussions to form collective decision-making recommendations. Similar experiments also exist in Web3. However, to truly test the feasibility of random sampling, these citizen assemblies need to have a certain degree of substantive decision-making power, rather than just making suggestions, and observe the effectiveness of their resolutions. However, for many significant issues, we need representatives or token holders to conduct in-depth research and continue iterating, as such topics are not suitable for 'one-off' sampling meetings (because once the meeting is completed, it disbands). Next year, we will also observe how far random sampling can go.
Author's Biography
Andrew Hall is a professor of political economy at Stanford University's Graduate School of Business and Political Science. He works in the a16z research lab and serves as an advisor to technology companies, startups, and blockchain protocols, primarily studying the intersection of technology, governance, and society.