March Session: Applied Intelligence: INTEGRATING AI TECHNOLOGIES INTO HEALTH PROFESSIONS EDUCATION-Applied Intelligence: INTEGRATING AI TECHNOLOGIES INTO HEALTH PROFESSIONS EDUCATION

Video Transcription

Video Summary

The webinar opened with Dr. Miguel Paniagua introducing the CMSS Artificial Intelligence/Machine Learning Task Force and framing the session around AI’s growing impact on medical societies and health professions education. The main speaker, Dr. Laura Turner, then presented “Applied Intelligence: Integrating AI Technologies into Health Professions Education.”<br /><br />Dr. Turner argued that AI is already deeply embedded in both clinical care and education, but its strengths and limits must be understood carefully. She explained that large language models are probabilistic pattern predictors, not true reasoners, which makes them prone to “hallucinations” and especially dangerous when they omit critical information. She introduced the idea of the “alignment paradox” and described a one-over-x learning curve: AI performs best in routine, high-frequency tasks, but humans remain essential for complex, uncertain, and rare clinical situations.<br /><br />She reviewed evidence showing AI can grade notes, evaluate encounters, answer patient questions empathetically, and even outperform physicians on some benchmarked cases, yet still fails on adaptive reasoning and safety-critical omissions. She highlighted concerns about cognitive debt and skill atrophy when learners over-rely on AI.<br /><br />Dr. Turner distinguished between training and education, arguing that AI is well suited to training routine skills, while education must preserve human judgment and uncertainty management. She showcased her lab’s AI-based simulation platform, which provides standardized patient encounters, personalized feedback, and real-time assessment of reasoning patterns. She concluded that AI should augment, not replace, human coaching and could help solve Bloom’s two-sigma problem by enabling scalable one-on-one tutoring.

Keywords

artificial intelligence

machine learning

health professions education

large language models

clinical simulation

cognitive debt

skill atrophy

medical societies

AI tutoring

human judgment