Large Language Models -- like the ubiquitous ChatGPT -- have taken the world by storm. This talk is *not* about ChatGPT but rather about Large Language Models themselves, with a particular focus on open source LLMs and whether they can be taught to create metadata.