DAOs rely on an engaged community to vote on proposals thoughtfully. Every proposal needs to be voted on manually which slows things down and leads to DAO fatigue. Source: Alex Cheema & Mohamed Baioumy from AI x Crypto Primer

Automate proposals by giving members of a DAO an AI Agent that votes on their behalf based on their values. The model could use a Bayesian approach which takes into account the uncertainty in the decision being made. If the uncertainty is above a threshold, then bring the proposal to the attention of the members of the DAO and require a manual vote. The onboarding process for creating these AI agents needs to be streamlined for users to adopt this. A combination of on-chain and off-chain data from each user can be used to define the preferences of a user without too much manual work.

Source: Alex Cheema & Mohamed Baioumy from AI x Crypto Primer Full credits go toward these two legends. I wanted to share some incredible idea from their presentation & primer.

pragma solidity ^0.8.24; contract Example { function decide(string memory proposal) public pure returns (Choice, string memory) { string memory result = llmInference(string.concat(prompt, proposal)); string memory v = substring(result, 0, 1); string memory reason = substring(result, 2, bytes(result).length); if (keccak256(abi.encodePacked((v))) == keccak256(abi.encodePacked(("Y")))) { return (Choice.For, reason); } else if (keccak256(abi.encodePacked((v))) == keccak256(abi.encodePacked(("N")))) { return (Choice.Against, reason); ^some example code, unfinished