Grey’s AI-natomy: How to Effectively Use ChatGPT in Medical Education

ChatGPT and other Large Language Models (LLMs) can improve educator and student performance in curriculum design and patient interaction simulations. Success relies on well-crafted prompts that effectively communicate tasks and context to the AI. This workshop will focus on teaching prompt crafting skills for enhanced medical education and avoiding AI errors and “hallucinations.” Interactive activities will be included, and ready-to-implement examples will be shared with participants.

Learning objectives:

• Apply best practices in “prompt engineering” to effectively use Large Language Models (LLMs) like ChatGPT in medical education for the creation and delivery of educational content, as well as for the assessment of student learning.
• Critically analyze the output of LLMs, distinguishing between accurate information, AI hallucinations, and potential biases in the AI outputs.
• Recognize the limitations and capabilities of LLMs, as well as the role of the quality of the initial prompt in determining the quality of the output.
• Synthesize knowledge about ethical considerations, such as data privacy and broader implications of relying on AI for education, in the use of LLMs in medical education.