PIG
Trust Paradigm Shift: An Epistemic Trust Protocol in AI
This proposal introduces an epistemic trust protocol aimed at reshaping how trust is established and managed in AI systems within decentralized automated organizations (DAOs). By focusing on a decentralized, transparent, and dynamic approach, this project addresses the limitations of centralized trustworthy AI protocols that typically benefit large, powerful organizations. The proposed protocol seeks to allow individuals and communities to develop trust in AI based on technical and interpersonal dynamics facilitated by effective oversight. This paradigm shift advocates for recognizing and integrating individual and diverse preferences in the age of AI, promoting a more participative and adaptable trust framework in digital environments.
Author: Xiaoou He, Xiaomeng Hu