AI Scribes in Med Ed: Why We Shared Yale’s Latest Insights
.png)
As artificial intelligence rapidly transforms clinical workflows, medical educators face a unique dilemma: How do we prepare trainees for an AI-driven future without undermining their core diagnostic skills? A recent piece from the Yale School of Medicine caught our eye because it directly explores this double-edged sword. Written by Serena Crawford, Associate Communications Director at Yale School of Medicine, the Q&A highlights the pioneering work of Jaideep Talwalkar, MD, Associate Dean for Educational Technology and Innovation, as he examines how ambient AI scribes impact medical students.
| Key Takeaways |
| Guardrails Matter: Yale's approach requires students to write their own initial note first, using the AI draft only as a secondary reflection tool to build a "hybrid" note without sacrificing critical thinking. |
| Preserving Clinical Reasoning: To prevent "never-skilling," Dr. Talwalkar’s team is building custom educational AI tools that deliberately suppress the automated assessment and plan sections, forcing trainees to work through diagnosis independently. |
| Enhancing Human Connection: When implemented thoughtfully, ambient tools remove the barrier of the computer screen, allowing trainees to be fully present with their patients. |
Connecting Clinical Research to Educational Practice
What makes Dr. Talwalkar’s commentary particularly compelling is that it grounds educational policy in rigorous empirical evidence. His insights directly mirror the findings published in our journal, JMIR Medical Education, in the paper "Impact of an Ambient AI Scribe on Medical Student Objective Structured Clinical Examination Notes: Nonrandomized Clinical Trial."
In his discussion, Dr. Talwalkar points out a crucial risk: when medical students are given access to ambient AI scribe outputs before synthesizing their own clinical reasoning, the quality of their assessment and plan sections can degrade. AI assistance can inadvertently act as a crutch, leading to early deskilling if guardrails aren't intentionally built into training.
At JMIR Publications, we believe that sharing these real-world institutional perspectives alongside peer-reviewed research helps bridge the gap between study findings and practical classroom implementation.
Read the full Q&A from Yale School of Medicine to explore Dr. Talwalkar’s full recommendations for integrating AI into health professions education.