From bugs to bots: learning infectious diseases in the AI era

As a millennial, I have experienced the shift from an analog world to a digital universe.This transition has shaped many parts of my life, including my education.My journey in education began with government-issued textbooks during elementary school. 1 It continued through the explosion of the internet and the World Wide Web, which fueled the rapid evolution of digital reference tools, such as encyclopedia Encarta and later Wikipedia.As digital connections grew faster, new tools emerged in the palm of my hand through smartphones; from digital books to podcasts and YouTube videos.However, as these learning resources become increasingly accessible, the sheer volume of information often exceeds what one can reasonably review or study.Furthermore, the more specialized a medical field becomes, the more clinicians must rely on lengthy guidelines, reviews, and research papers, many of which are updated far more rapidly than other sources.Now, a new tool has captured my attention and addresses this challenge.It even helped me prepare more efficiently for my Infectious Diseases board examination: large language models (LLMs).LLMs deployed as conversational chatbot tools, such as ChatGPT, have captivated the world, including the medical community.They have demonstrated strong performance on standardized tests, the ability to generate humanlike responses, and empathetic communication. 2 This has led to the anthropomorphization of chatbot tools, overshadowing tasks that LLMs truly excel at, such as data summarization, in the push to replicate human clinical reasoning.In fact, up to 82% of recent studies on LLMs in medicine have focused on questionanswering, whereas only about 9% have examined summarization tasks.