Educating the Next Generation: Making Large Language Models Mandatory in Medical Training

The deployment of Large Language models (LLMs) introduces significant opportunities in the medical domain. Starting from supporting medical documentation to clinical decision making, the applications appear boundless. However, this brings challenges such dissemination of existing societal biases or the generation of misinformation. To address a positive future for LLMs in medicine, balancing both opportunities and challenges, we have designed and implemented a new mandatory course for medical students as part of their core curriculum. It includes a 90-minute session divided into three parts: a technical overview of LLMs, a live Turing Test to experience and evaluate human versus machine-generated responses, and a practical group exercise to assess the application of LLMs in three different medical tasks. This paper discusses the course structure, release the implementation of our Turing Test and LLM communication interface, and provides an initial evaluation, illustrating the critical role of education and fostering responsible use of emerging LLM technology in healthcare and research.