Normative Concept

Normative concepts, encompassing ethical principles and societal rules, are increasingly central to artificial intelligence research, particularly concerning the development and deployment of large language models and other generative AI systems. Current research focuses on identifying and mitigating biases, harmful outputs, and other ethical concerns within these models, often employing techniques like reinforcement learning from human feedback and frame-based approaches to construct and apply norm bases. This work is crucial for ensuring responsible AI development, addressing issues of fairness, safety, and societal impact, and ultimately shaping the future of human-AI interaction.