Projected hetero-conciliation

how humans sustain communication with generative AI without shared intentionality

Abstract

This article examines the cognitive conditions that facilitate communication between humans and generative artificial intelligence systems, focusing on the inferential architecture that underlies communicative action. Although the artificial agent lacks intentionality, interaction achieves functional coherence through the heuristic projection of collaborative agency by the human participant. To account for this phenomenon, the study introduces the concept of projected hetero-conciliation, defined as the provisional attribution of communicative intention to the system to maintain continuity of the action plan. Employing a theoretical-analytical approach, the research compares human–human and human–generative AI exchanges to evaluate the descriptive adequacy of pragmatic-cognitive models. Findings demonstrate that communication persists through human inferential mechanisms, underscoring the need to extend the theoretical framework to contexts devoid of reciprocal mental states.