DeFi DAOs claim to have decentralized governance but have a problem of governance tokens concentrated in the hands of developers, investors and whales.
Achieving true decentralization and plurality requires going beyond quadratic voting mechanisms by measuring the uniqueness of each voter. For this project, a new DAO voting mechanism is made be creating voting uniqueness scores, which is based on the type of proposal as determined by a Machine Learning clustering algorithm.
This project uses machine learning to assign each voter a uniqueness score. Filecoin IPFS allows us to store the input data, intermediate ML model outputs and uniqueness scores assigned to each user for each DAO proposal vote with provable authenticity and Chainlink oracles external adapter allow us to read the data using a smart contract.
Voter uniqueness is based on the type of proposal as determined by a Machine Learning clustering algorithm. Voter weights (also termed uniqueness scores) range from 0 to 100, and are calculated based on the voter’s composition of voting patterns.