Implementing Large Language Models in Critical Care

Abstract

Large language models are artificial intelligence tools that have emerged in a supportive role in various fields, and we are on the cusp of witnessing their true potential in critical care medicine. In this article, we highlight the anticipated applications of these models in this setting, as well as the barriers that separate us from actualizing their deployment, ranging from pragmatic concerns, such as doubts in clinical decision-making, to ethical trustworthiness, along with means to attempt to mitigate these challenges.