Deep Tide TechFlow News, February 23 — According to Cointelegraph, Vitalik posted on X platform on Sunday proposing to introduce personal AI large language models (LLMs) into decentralized autonomous organization (DAO) governance to address the long-standing issue of low participation rates in DAOs. Currently, the average voting participation rate in DAOs is only between 15% and 25%, which not only risks centralization of power but also exposes governance to attack risks.
Vitalik pointed out that the core challenge of democratized and decentralized governance is “the limited attention span of humans,” and existing delegated voting mechanisms tend to allow a small number of representatives to control decision-making, while other members almost lose their voice. He suggested that personal AI assistants could infer user preferences based on writing records, conversation history, and direct statements, and then cast votes on their behalf; if there is uncertainty about an important issue, the AI should proactively ask the user and provide relevant background information.
Regarding privacy protection, Vitalik proposed that personal LLMs could be placed in a “black box” environment to handle sensitive information, only outputting the final decision result, thus supporting governance decisions involving private information while safeguarding privacy. Lane Rettig, a researcher at the Near Foundation, stated that the organization began exploring similar solutions last year, such as using AI digital avatars to vote on behalf of DAO members.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Vitalik proposes introducing AI-assisted DAO governance to address the issue of low voter participation.
Deep Tide TechFlow News, February 23 — According to Cointelegraph, Vitalik posted on X platform on Sunday proposing to introduce personal AI large language models (LLMs) into decentralized autonomous organization (DAO) governance to address the long-standing issue of low participation rates in DAOs. Currently, the average voting participation rate in DAOs is only between 15% and 25%, which not only risks centralization of power but also exposes governance to attack risks.
Vitalik pointed out that the core challenge of democratized and decentralized governance is “the limited attention span of humans,” and existing delegated voting mechanisms tend to allow a small number of representatives to control decision-making, while other members almost lose their voice. He suggested that personal AI assistants could infer user preferences based on writing records, conversation history, and direct statements, and then cast votes on their behalf; if there is uncertainty about an important issue, the AI should proactively ask the user and provide relevant background information.
Regarding privacy protection, Vitalik proposed that personal LLMs could be placed in a “black box” environment to handle sensitive information, only outputting the final decision result, thus supporting governance decisions involving private information while safeguarding privacy. Lane Rettig, a researcher at the Near Foundation, stated that the organization began exploring similar solutions last year, such as using AI digital avatars to vote on behalf of DAO members.