Large language models (LLMs) have profoundly changed clinical natural language processing (NLP) and brought substantial opportunities and challenges for artificial intelligence (AI) in healthcare. LLMs almost become a unified solution for clinical NLP, yet current findings are mixed with prevails and mirages. On one side, LLMs have been reported to be able to pass several professional exams, yet there is limited evidence supporting the benefit of LLMs to many classic clinical NLP tasks.
This talk will introduce the technologies powering LLMs, overview the recent prevails, and examine the mirages in the hype of LLM magic. Based on the experience in developing two clinical LLMs in the clinical domain, including GatorTron and GatorTronGPT, this talk will provide insight into the potential application of LLMs for clinical NLP and healthcare.