Annie Yuan's Tacit Signal Infrastructure: AI that learns expert intuition over time

Annie Yuan's vision paper, Tacit Signal Infrastructure: Towards AI Systems that Model Expert Sensing Over Time (arXiv:2605.24332), argues that current AI systems excel at processing explicit knowledge—retrieving, summarizing, generating explanations—but fail to capture the tacit layer of expert cognition. Experts rely on subtle sensing: perceiving weak signals, recognizing emerging tensions, detecting degradation in coherence, and anticipating instability before formal metrics show anything. The paper calls for next-generation AI to move beyond prompt-based productivity tools toward systems that model this longitudinal expertise.

Yuan defines Tacit Signal Infrastructure as a dedicated layer for capturing, structuring, modeling, interpreting, and validating expert tacit signals over time. She introduces Long-term Cognitive Operations, which include memory curation, semantic organization, tacit signal modeling, reasoning calibration, and cognitive governance. To operationalize this, she proposes the Cognitive Operations Manager—a new AI-native professional role that coordinates these tasks. Finally, the Cognitive Operations Research and Training Framework (CORTF) supports research, education, and workforce development. The 17-page paper, including two figures, positions cognition as an infrastructural, operational, and professional domain for persistent human-AI systems.