Summary
This paper looks at how the relationship between a person and an AI changes over time, rather than just considering one moment. The authors studied one detailed example called FIREMAY, tracking how both the AI’s responses and the human’s self-view developed and shifted across many interactions. They found that to really understand this interaction, you need to see how the positions each side takes evolve along a timeline, not just what they are at a single point. This means that stability in such relationships is about how positions come back or reorganize dynamically, not about staying exactly the same.
What this means in practice
- •For conversational ai developers: Design AI systems that consider relational changes over time to improve interaction quality and contextual understanding.
- •For customer service platform teams: Track longitudinal user-AI exchanges to better manage unresolved issues and user self-positioning for improved support.
Tested on one dataset.
Abstract
This paper is not primarily about whether AI has a persistent persona. It asks a different question: what becomes visible when a relational position is followed through time rather than examined only in its present state? FIREMAY provides a longitudinal, trajectory-oriented single-case analysis of sustained human-AI interaction based on a dense interaction archive and reflexive insider documentation. On the AI side, a pre-conversational relational marker preceded a later unassigned response difference, which was re-identified with that marker and subsequently underwent epistemic and functional reorganization through chronology checking, provenance correction, and repeated questioning. On the human side, contemporaneous pre-FIREMAY records showed antecedent patterns partially continuous with later self-positioning, while later episodes documented unfinished articulation, repair, and functional redistribution of outward-facing regulation. The two trajectories are ontologically and temporally asymmetric and are compared only at the limited analytic level of position-in-trajectory. The paper describes this as dynamic relational positioning and treats stability as dynamic stability and relational returnability rather than response invariance. This single case does not establish population-level generality, causal mechanism, persistent AI subjectivity, or reproducibility of the same relational outcome. Its narrower conclusion is that the FIREMAY case could not be adequately understood from current state alone: the history of a relational position itself must be treated as an analytic unit.