Transition‑Aware Operating State Model for Life‑Centric AI

Intelligent user interfaces typically assume a stable user context. They personalise gradually, optimise for productivity, and maintain consistent interaction styles. However, human life is punctuated by transitions—job changes, relocation, illness, divorce, bereavement—during which user needs, availability, and risk tolerance change dramatically. Standard systems, unaware of the transition, continue to send notifications, suggest optimisations, and request feedback—behaviours that become intrusive, irrelevant, or harmful.This paper introduces the Transition-Aware Operating State Model (TAOSM)—a finite state machine that treats life instability as a first-class operating state. The model comprises six states: STABLE, SENSING, EMERGING, CRISIS, RECOVERY, NEW_STABLE. We present the Seismic Detector, a continuous monitoring system that analyses typing dynamics, communication tone, calendar volatility, keywords, and interaction withdrawal to compute a stability index (0–100) that drives state transitions.Evaluation through a simulated scenario test suite (N=100 logs) achieves 87% sensitivity and 4% false positive rate for transition detection. A user study (N=30) comparing standard vs. transition-aware system behaviour during simulated life events shows statistically significant improvements in appropriateness (4.7 vs. 2.1/5), trust (4.3 vs. 2.5/5), and perceived empathy (4.8 vs. 1.4/5), p < 0.001 for all metrics. This is Paper 2 of 8 in the constitutional Identity Series from Nexus Sovereign. Targets IUI 2027 (Intelligent User Interfaces conference) This research is conducted under the Nexus Sovereign Research imprint and implemented commercially through Pietarien (https://pietarien.com), the company building the Nexus Business and Personal operating systems.