TIL that large language models (LLMs) work by taking an input text and repeatedly predicting the next token or word. Most notable models include the GPT series.

https://en.wikipedia.org/wiki/Large_language_model

https://www.reddit.com/r/todayilearned/comments/15mlghf/til_that_large_language_models_llms_work_by/