Quick Buzz Feed
Robert Wachter, MD, chair of the Department of Medicine at UC San Francisco and author of a new book on AI and health care, discusses AI’s transformative power.
Medical schools must identify new competencies for the AI age, focusing on comfort with AI tools, understanding their capabilities, and adapting to change. Simultaneously, they need to re-evaluate and potentially remove legacy competencies that are no longer essential, such as extensive memorization of drug mechanisms or the Krebs cycle, while retaining critical skills like diagnostic reasoning. The goal is to prepare future physicians for a rapidly evolving healthcare landscape where AI plays a significant role.
AI offers valuable opportunities for student coaching. For instance, it can help students prepare for sensitive patient interactions, such as discussing end-of-life care. By simulating scenarios and providing feedback, AI can create a safe environment for students to practice and refine their communication skills before facing real-world situations, addressing anxieties associated with such critical conversations.
AI can significantly improve educational assessment, particularly for clinical skills. Currently, evaluating interviewing skills often relies on imperfect proxy measures. With AI, a student's conversation with a patient (with permission) could be recorded and analyzed by the AI, providing precise, detailed feedback on their technique. This represents a major advancement over traditional methods, offering comprehensive insights that human observers might miss due to time constraints.
Students should avoid or postpone the use of AI for tasks that require developing foundational critical thinking. Premature reliance on AI tools for activities like note summarization or generating differential diagnoses can bypass crucial 'educational friction' necessary for developing neural pathways for independent thought. Students need to master these skills unaided before integrating AI, to prevent 'de-skilling' or 'never-skilling' where they become overly dependent on technology to outsource their thinking.
To mitigate the risks of AI, tools should be designed to encourage active learning and critical engagement. Instead of providing immediate answers, AI should prompt trainees to formulate their own responses first, then offer constructive critique. This approach ensures students remain engaged in the learning process and develop foundational skills. Professionals must be proficient in their core competencies without AI before effectively partnering with the technology.
Teachers face the complex task of assessing students' competencies in a rapidly changing environment. They need to ascertain what students can accomplish independently, without AI, before gradually introducing AI as a 'copilot.' Furthermore, educators must justify teaching content that AI can readily provide, often needing to emphasize the importance of foundational knowledge, even if it feels 'old school' to students, to ensure a robust understanding.
There is a significant risk of 'never-skilling' or 'de-skilling' if medical students become excessively reliant on AI. To counter this, strategies like periodically deactivating AI tools or programming them to require a clinician's initial medical opinion before offering a solution may be necessary. This approach ensures that students and practicing physicians maintain and actively use their foundational skills, preventing a scenario where critical abilities are lost due to over-automation.
Despite the prevalence of AI in crafting and assessing application essays, the essay retains its utility in revealing a student's priorities and communication style. The focus shifts from the writing quality itself to the substance of what the student chooses to convey, even if AI-assisted. Given the rapid changes, medical education must continuously re-evaluate traditional methods, acknowledging that few practices from the past century can remain static.