While artificial intelligence has advanced in perception, planning, and simulation, it remains limited in its capacity for social intelligence; particularly in engaging with the patterned, tacit, and emergent dimensions of human interaction. This paper introduces a socio-cultural framework for Artificial Social Intelligence (ASI), grounded in Edward T. Hall's theory (1959) of intercultural nonverbal systems. Rather than viewing social behavior as a set of discrete cues, the model conceptualizes Hall's systems- such as association, territoriality, learning, and play- as emergent, context-sensitive dimensions of cultural interaction. Hall's matrix is reinterpreted as a multidimensional interaction space, enabling artificial agents to recognize informal norms, adapt dynamically, and align through participation in ongoing interaction. By bridging socio-cultural theory with developments in Human-Robot Interaction and emerging approaches to Theory of Mind and self-adaptive architectures, the paper advances a theoretical foundation for designing AI systems that participate meaningfully in human social interaction.
Anat Ringel (Tue,) studied this question.