The accelerating integration of artificial intelligence (AI) into clinical practice presents a profound challenge to conventional medical education and assessment. Legacy assessment methods, focused on individual knowledge recall, fail to evaluate the competencies required for effective human-AI partnership. While numerous AI competency frameworks have emerged, they often overlook the synergistic wisdom needed for true collaboration. To address this gap, this article introduces and defines collaborative intelligence (CI), a dyad-centric construct focusing on the emergent capabilities of the human-AI team. CI represents the capability to critically appraise, ethically integrate, and judiciously act upon AI-derived insights. Its 5 core components are delineated: critical AI appraisal, information synthesis, adaptive judgment, ethical reasoning, and effective human-AI interaction. Subsequently, it proposes a novel assessment framework that discards rigid, time-based training stages in favor of a 4-stage competency spiral (novice to expert). This programmatic approach systematically maps assessment methods-including workplace-based assessments, high-fidelity simulations, and script concordance testing-to each CI component across the developmental spiral. Reengineering assessment to effectively measure CI is presented not merely as a technical adjustment but as a strategic imperative for cultivating physicians who possess the wisdom to harness algorithmic power while upholding the humanistic core of medicine, thereby ensuring safe and equitable care in the coming century.
Yanyi Wu (Wed,) studied this question.