The Near Foundation is innovating by creating AI-driven “delegates” to improve governance vote participation among its decentralized autonomous organization (DAO) members.
As the rollout of these AI delegates progresses, it will be executed in stages, beginning with chatbot-like models and eventually advancing to individual digital twins for every DAO member.
Lane Rettig, a researcher at Near Foundation specializing in AI and governance, mentioned that this overhaul is still under development, underlining the Foundation’s oversight of the layer-1 Near Protocol.
The goal is for each user’s digital twin to learn their preferences for governance matters, thus enabling faster decision-making akin to a “math problem” that can be solved almost instantly. Rettig expressed during an interview at the Token2049 conference in Singapore, “Then you kind of set this thing loose, and it kind of acts on your behalf and votes on your behalf.”
“That’s in a sense, almost our end game vision for this, where we replace all human actors with a digital twin, if you want to call it that, to solve this voter apathy, participation issue.” Translation: “This is nearly our ultimate vision for replacing human participants with a digital counterpart to tackle the problem of low voter engagement.”
Currently, average participation in DAOs tends to hover between 15% and 25%. This low rate can lead to centralization of authority and poor decision-making, posing risks like governance attacks where malicious actors utilize excessive tokens for harmful proposals without detection.
Human Involvement is Essential
Rettig emphasized the necessity of a human element within this governance framework, stating, “I think there’s definitely a category of things where you’re going to want the human to make the final decision, pull the trigger.”
He elaborated that AI, while a useful tool, should support rather than replace human decision-making.
“And having said that, it can not only nudge you, it can also say, based on what I know about you, I think you should vote this way, but you should be the one to vote right, and they can learn.” Translation: “While it can suggest optimal voting choices based on gathered data, the human should ultimately have the casting vote.”
AI entities are already frequently used throughout the crypto space to enhance Web3 applications autonomously, manage service interactions, and even trade cryptocurrencies.
Training Delegates by User Behavior
Just like generative AI chatbots, the AI delegates will be trained through user interactions such as interview sessions, voting history analysis, and messaging on platforms like Telegram and Discord.
Rettig noted that the number of AI agents in the crypto sector is projected to exceed 1 million by 2025, based on estimates from investment manager VanEck, which highlights ongoing concerns related to the potential security vulnerabilities they may introduce.
To address these issues, Near Foundation is focusing on a verifiable training model that ensures the delegates adhere to users’ values.
Gradual Deployment of Delegates
Rettig explained that the initial deployment of AI delegates would involve simpler applications with limited agency, functioning similarly to chatbots. The scaling process will initially assign delegates for large voting groups before transitioning to individual representatives for each voter.
“Then governance becomes a math problem, you’re just summing it up. Every time a vote comes up, it can happen almost instantly because you have all the agents there, and they know how everyone will vote ahead of time, then boom, you’re done.” Translation: “This approach turns governance into a numerical calculation, leading to rapid outcomes since agents will already forecast voting patterns.”
