A Multi-Level Framework for the AI Alignment ProblemDownload PDF

Published: 05 Dec 2022, Last Modified: 05 May 2023MLSW2022Readers: Everyone
Abstract: AI alignment considers how we can encode AI systems in a way that is compatible with human values. The normative side of this problem asks what moral values or principles, if any, we should encode in AI. To this end, we present a framework to consider the question at four levels: Individual, Organizational, National, and Global. We aim to illustrate how AI alignment is made up of value alignment problems at each of these levels, where values at each level affect the others and effects can flow in either direction. We outline key questions and considerations of each level and demonstrate an application of this framework to the topic of AI content moderation.
1 Reply