Whilst typical AI user interfaces are designed to facilitate the illusion of a person-to-person interaction, the user is, of course, engaging with a vast ‘social’ network distributed across human, non-biological and digital nodes, enabled by an infrastructure comprising human intellectual labour, data centres, energy resources etc. How should this sociality be modelled? What is its ontological status? The participation in social networks that enables human intelligence is reflected in the very nature of that intelligence – but does this, as is often claimed, provide a model for understanding the relation between AI and its (very different) enabling conditions? The Kyoto workshop will be devoted to addressing these and related questions.