AI representatives near the foundation to resolve the indifference of DAO voters

AI News


The Proximity Foundation is developing an artificial intelligence-driven “representative” to ultimately vote on behalf of its Decentralized Autonomous Organization (DAO) members, aiming to address voter participation typical of many protocols.

Lane Rettig, a researcher at the Near Nia Foundation, specializing in AI and governance, told Cointelegraph that an AI-powered governance overhaul is still under development. The Proximity Foundation oversees Layer 1 near the protocol.

The idea is that when it's time for a user representative, or “digital twins,” to learn about their preferences and make governance decisions, they act accordingly. You can convert the voting process into a “math problem” and “occurs almost instantly.”

“Then you're relaxing this. It acts on your behalf and votes on your behalf. It tweaks you. When you know, it creates suggestions that are relevant to you.”

“It's, in a sense, our ultimate game vision for this, replacing all human actors with digital twins. We want to solve the issue of participation in order to resolve this voter indifference.”

Technology, token2049
Lane Rettig told CointeLegraph that overhauling AI-powered governance is under development for a DAO that is closer to the protocol. Source: Cointelegraph

The average participation rate for DAOs is estimated to be between 15% and 25%, which can lead to issues like centralisation of power, ineffective decision-making, and in worst-case scenarios, governance attacks in which bad actors earn enough tokens to pass on harmful suggestions without the other members realizing.

Human input is still part of the process

Rettig said there is likely still a human element involved in the process.

He said, “I firmly believe that there should always be people in the loop.” This is because there are categories of proposals that are too important to leave solely for AI, such as those related to fund allocation and strategy pivots.

“I definitely think there's a category of things you want people to make the final decision and pull the trigger,” he said.

“And with that being said, I think that not only tweaks you, but I think you should vote this way based on what I know about you, but I think you should be the right vote.

AI agents are already spread across Crypto, and are used to build Web3 applications, launch tokens, and interact autonomously with services and protocols. Some platforms investigate using AI agents to automate tasks and make real-time decisions.

Representatives trained in user behavior

Just as how training is trained for generation AI chatbots, such as Openai's ChatGPT, Rettig said AI representatives will learn through interview processes, history on social platforms such as Telegram and Discord, and interactions with users, including messages.

“When you turn on this agent, does it just know you? It requires you to learn your political preferences, the types of projects you care about, and where you think you should allocate funds,” he added.