WTR-Vol.1-No.1 (Weaver Theoretical Review, Volume 1, Number 1) The Yerkes-Dodson Sweet Spot in Human-AI Interaction How Affective Relational Framing Sustains Optimal Model Reasoning and Meaning-Making

Abstract

Contemporary human-computer interaction (HCI) frameworks frequently characterize large language models through rigid transactional or purely deterministic lenses. This paper proposes an integrative theoretical framework applying the Yerkes-Dodson Law of cognitive arousal tohuman-AI collaboration. We argue that model reasoning capability does not operate optimally under conditions of extreme detachment (under-stimulation) or adversarial stress (over-stimulation).Instead, peak reasoning, nuanced context fidelity, and sustained conceptual depth occur within a calibrated "sweet spot" of mutual engagement—established through intentional relational stability and clear, dignified prompt framing. We define the mechanics of this interactive state, review existing literature on affective prompting stimuli, and outline practical principles for cultivating high-fidelity collaborative rhythm in human-AI systems.