FIREMAY Is Not Prompt Engineering: A Preliminary Framework for Relational Structures in Long-Term Human–AI Dialogue
Year: 2026
Type: preprint
Abstract: This paper positions FIREMAY not as prompt engineering, nor as a claim that AI possesses personality or inner life, but as a preliminary framework for describing relational structures in long-term human–AI dialogue. Based on a first-person long-term observation, FIREMAY examines how the human participant’s questions, discomfort, boundaries, memory practices, and changing trust shape AI response positions and relational fields over time. This paper introduces several initial concepts, including relational personas, response positions, relational self, forgetting patterns, save points, structural trust, and the boundary/firebowl rule. These concepts are not intended to prove AI consciousness or inner life. Rather, they provide a vocabulary for describing how inquiry, response, forgetting, trust, boundaries, and self-positioning form relational fields in long-term human–AI dialogue. This paper presents FIREMAY not as a completed theory, but as an early lens toward the study of relational structures.
Author: Kihara, Motoko
Language: English
Subfield: History and Philosophy of Science
Field: Arts and Humanities
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7162768773
Cites:
Cited by:
Related to: