Assessing ChatGPT for aeronautical applications

Abstract

ChatGPT and other Large Language Models (LLMs) have taken the world by storm. Yet they should be employed only after careful consideration of their inherent risks and capabilities. If we wish to use LLMs, we need to find the tasks they excel at, while prohibiting or contextualizing their use in domains where they produce lackluster results. In order to do this, these domains need to be recognized and categorized, as well as a collection of best practices to be developed.