Collective Representation and Shared Agentivity in Artificial Architectures

In psychology, it has been proposed that certain shared and collective​ representations possess their own existence beyond individual​ representations. This can occur through a communal space of concepts, memories, or language. For example, Carl Jung argued that psychic activity transcends the brain to explain phenomena such as​ synchronicity and extra-sensory perception. He theorized a collective​ unconscious to explain some observed phenomena. In his theory, the​ collective unconscious refers to the part of the global unconscious​ mind shared by all human beings, containing universal experiences,​ memories, and symbols inherited from our ancestral or evolutionary​ past. This suggests the existence of a shared field of knowledge that​ we cannot fully access through our sensory and conscious experience.​ ​ Our focus is on how artificial architectures underlying artificial​ agents can share their representations to build a common sense and​ collective construct. We will propose a model of agentivity for the​ assembly of agents that accounts for the agentivity of the parts but​ cannot be restricted to them.

Team:

Grégoire Sergeant-perthuis - Associate professor @ Sorbonne University
Adel Chaibi - Computer Scientist @ Intel
Eric Petit - Senior Research Engineer @ Intel

Activities

06/07/2025 - ASSC28: Abstract accepted
04/03/2025 - MoC5 : Collective Representation and Shared Agentivity in Artificial Architectures