Magical Governance
TLDR: 120 ETH to support the development of Wand. A project aiming to scale onchain governance through the use of automated help or "agents".
Overview
Over the past several years, Nouns has made significant strides in governance and has served as a shining example of how a decentralized community can operate. However, the inherent complexities of decentralization have made it challenging to retain existing participants and attract new ones.
As our community continues to evolve, we face challenges like members leaving and an increase in information asymmetry. In addition to these challenges, the demands of governance during times of rapid growth make voting a challenge for many.
Fortunately, recent advancements in LLMs offer us a new set of tools to govern more efficiently. Much of governance involves interacting with others and processing information using natural language—tasks that LLMs excel at. Plus, our culture of openness and public feedback has uniquely positioned us to get more out of LLMs, as we generate a wealth of data through onchain governance (vote history, reasons, preferences, etc.).
Approach
Wand is building governance-aware AI. AI that understands us, our values, and augments our ability to govern well. For Wand to become a valuable tool in governance, the following capabilities are crucial:
- Analyze voter activity: Employ vote simulation and predictions to reduce the cognitive burden associated with governance participation.
- Index and surface relevant information: Provide easy access to voter, proposal, and other related information, enabling the community to make more informed decisions.
- Automate proposal review and analysis: Offer automated feedback on proposals to enhance their quality before submission.
By enabling LLMs to understand and participate in governance, we can attract and retain more participants, ultimately increasing the likelihood of success for Nouns.
Prototype / Demo
Visit https://wand.wtf and you'll see automatically generated vote predictions for pending proposals and recent candidates. These predictions are based on each user's voting history and governance activity. They provide insight into what people care about, a valuable resource for other members and proposers seeking feedback on their ideas.
I anticipate the quality and accuracy of these predictions will increase significantly with further development. The current implementation is a work in progress but serves as a compelling demonstration of an agentic workflow that can help with decision making.
Ideally, Wand will evolve to a state where it can handle the majority of routine governance tasks. Wand does not aim to replace human voting but rather to enable Nouns to scale its governance participation significantly (although I envision a future where many choose to leverage Wand to automate most of their activity).
Goals / Roadmap
Short term goals for Wand:
- Enhance vote predictions and reasoning pipeline: This is where most of the magic happens
- Implement customizable user preferences and actions: Allow voters to define their individual preferences and automate actions on Wand, such as automatically voting "yes" on proposals supporting specific causes.
- Develop automated proposal review and feedback tools: Provide tools for automatic proposal analysis, including quality checks, security, insights, etc
- Lower the barrier to entry for governance participation: Develop features that empower more individuals to become active and productive governors.
Some north star metrics are:
- Increase proposal throughput without increased governance fatigue (We need to be the best at allocating capital efficiently)
- Increased governance participation (Wand should empower a larger and more diverse group of individuals to feel confident and engaged in the governance process)
Ideally, Wand is useful for both community members and those seeking to open proposals. To continue to attract high quality talent to Nouns, it needs to be easier to open proposals or test community reception towards ideas.
Where the $ will go
The 120 ETH in funding will help support the development of Wand over the next 9 months. This covers expenses like:
- Infrastructure / servers
- AI Model inference / gpu time
- Team augmentation
- Research
All funds will be stored in a 3/5 multisig at address 0x6B81...D8e7.
Who am I
My contributions to the Nouns ecosystem over the past several years include:
- Developed Federation Governance Pools (vote market)
- Founded Plutocats (Nouns inspired decentralized community on Blast)
- Supported the development of a Nouns minting economy on Zora by building Bound
- Created a model for generating Nounish images
- Active participant in governance with a long voting history, as a Nouns gardener, and in multiple sub-DAOs in the ecosystem
- Creator of WizardsDAO (an early Nounish DAO)
Thanks for your time and consideration
Magical Governance
TLDR: 120 ETH to support the development of Wand. A project aiming to scale onchain governance through the use of automated help or "agents".
Overview
Over the past several years, Nouns has made significant strides in governance and has served as a shining example of how a decentralized community can operate. However, the inherent complexities of decentralization have made it challenging to retain existing participants and attract new ones.
As our community continues to evolve, we face challenges like members leaving and an increase in information asymmetry. In addition to these challenges, the demands of governance during times of rapid growth make voting a challenge for many.
Fortunately, recent advancements in LLMs offer us a new set of tools to govern more efficiently. Much of governance involves interacting with others and processing information using natural language—tasks that LLMs excel at. Plus, our culture of openness and public feedback has uniquely positioned us to get more out of LLMs, as we generate a wealth of data through onchain governance (vote history, reasons, preferences, etc.).
Approach
Wand is building governance-aware AI. AI that understands us, our values, and augments our ability to govern well. For Wand to become a valuable tool in governance, the following capabilities are crucial:
- Analyze voter activity: Employ vote simulation and predictions to reduce the cognitive burden associated with governance participation.
- Index and surface relevant information: Provide easy access to voter, proposal, and other related information, enabling the community to make more informed decisions.
- Automate proposal review and analysis: Offer automated feedback on proposals to enhance their quality before submission.
By enabling LLMs to understand and participate in governance, we can attract and retain more participants, ultimately increasing the likelihood of success for Nouns.
Prototype / Demo
Visit https://wand.wtf and you'll see automatically generated vote predictions for pending proposals and recent candidates. These predictions are based on each user's voting history and governance activity. They provide insight into what people care about, a valuable resource for other members and proposers seeking feedback on their ideas.
I anticipate the quality and accuracy of these predictions will increase significantly with further development. The current implementation is a work in progress but serves as a compelling demonstration of an agentic workflow that can help with decision making.
Ideally, Wand will evolve to a state where it can handle the majority of routine governance tasks. Wand does not aim to replace human voting but rather to enable Nouns to scale its governance participation significantly (although I envision a future where many choose to leverage Wand to automate most of their activity).
Goals / Roadmap
Short term goals for Wand:
- Enhance vote predictions and reasoning pipeline: This is where most of the magic happens
- Implement customizable user preferences and actions: Allow voters to define their individual preferences and automate actions on Wand, such as automatically voting "yes" on proposals supporting specific causes.
- Develop automated proposal review and feedback tools: Provide tools for automatic proposal analysis, including quality checks, security, insights, etc
- Lower the barrier to entry for governance participation: Develop features that empower more individuals to become active and productive governors.
Some north star metrics are:
- Increase proposal throughput without increased governance fatigue (We need to be the best at allocating capital efficiently)
- Increased governance participation (Wand should empower a larger and more diverse group of individuals to feel confident and engaged in the governance process)
Ideally, Wand is useful for both community members and those seeking to open proposals. To continue to attract high quality talent to Nouns, it needs to be easier to open proposals or test community reception towards ideas.
Where the $ will go
The 120 ETH in funding will help support the development of Wand over the next 9 months. This covers expenses like:
- Infrastructure / servers
- AI Model inference / gpu time
- Team augmentation
- Research
All funds will be stored in a 3/5 multisig at address 0x6B81...D8e7.
Who am I
My contributions to the Nouns ecosystem over the past several years include:
- Developed Federation Governance Pools (vote market)
- Founded Plutocats (Nouns inspired decentralized community on Blast)
- Supported the development of a Nouns minting economy on Zora by building Bound
- Created a model for generating Nounish images
- Active participant in governance with a long voting history, as a Nouns gardener, and in multiple sub-DAOs in the ecosystem
- Creator of WizardsDAO (an early Nounish DAO)
Thanks for your time and consideration