Presented at ICIS 2025, Nashville, USA
This paper formalizes when interaction-generated signals improve human-AI forecasting ensembles and validates an online fusion algorithm in simulation.
Preparing for journal submission
This design science study develops an auditable inference-time control layer that detects compassion deficits in LLM counseling responses and applies minimal targeted interventions.
Under review at Economics Letters
This paper establishes sharp identified sets for latent-class moments in selected-at-random positive-unlabeled data without valid instruments.
Presented at AOM Annual Meeting 2026, Philadelphia, USA (TIM Division)
This paper shows how allocations of AI access, improvement, exclusivity, and forking rights shape relationship-specific investment incentives and underinvestment across alternative governance regimes.
Working paper; available on SSRN
This paper models how dividing residual control over cross-client reuse and post-exit continuity shapes complementary investment in customer-specific AI learning assets across firm boundaries.
Published in Research Policy, 2026
This paper develops a game-theoretic model to analyze the pre-doctoral academic labor market dynamics in the age of AI.
Recommended citation: Wang, S. (2026). "Hope, signals, and silicon: A game-theoretic model of the pre-doctoral academic labor market in the age of AI." Research Policy, 55(8), 105568. https://doi.org/10.1016/j.respol.2026.105568.
Published:
This talk presents a theoretical framework and simulation study on optimizing business forecasting through human-AI decision fusion. The research explores how to effectively combine human judgment with AI predictions to improve forecasting accuracy and decision-making quality.
Graduate Course Workshop, J.Mack Robinson College of Business, Georgia State University, 2025
This course provides a comprehensive introduction to Agentic AI, exploring the principles, architectures, and applications of autonomous AI systems that can reason, plan, and act independently.