🤖 AI Summary
Current generative AI systems lack a dynamic, interpretable foundation in human values to assess behavioral “appropriateness,” rendering evaluations static and context-insensitive.
Method: We propose the first formal, cross-contextual and longitudinal theory of appropriateness, integrating sociological norm modeling, cognitive-neuroscientific hypotheses, and value-alignment analysis to construct a tri-layer dynamic judgment model—comprising social, cognitive, and technical levels.
Contribution/Results: Our work transcends conventional AI safety and ethics frameworks by systematically formalizing appropriateness as multi-scalar, context-dependent, and evolutionarily adaptive. It delivers an operational definition of appropriateness, interpretable value anchors grounded in human norms, and dynamically adaptable alignment pathways. These advances enable more robust, trustworthy, and human-aligned AI system design and governance, supporting rigorous, value-sensitive behavioral assessment across diverse temporal and situational domains.
📝 Abstract
What is appropriateness? Humans navigate a multi-scale mosaic of interlocking notions of what is appropriate for different situations. We act one way with our friends, another with our family, and yet another in the office. Likewise for AI, appropriate behavior for a comedy-writing assistant is not the same as appropriate behavior for a customer-service representative. What determines which actions are appropriate in which contexts? And what causes these standards to change over time? Since all judgments of AI appropriateness are ultimately made by humans, we need to understand how appropriateness guides human decision making in order to properly evaluate AI decision making and improve it. This paper presents a theory of appropriateness: how it functions in human society, how it may be implemented in the brain, and what it means for responsible deployment of generative AI technology.