Large language models (LLMs) have become the center of attention and hype because of their seemingly magical abilities to produce long stretches of coherent text, do things they weren’t trained on, and engage (to some extent) in topics of conversation