~/blog.html
LLMs are sophisticated functions that predict the next word (token) in a sequence by assigning probabilities to all possible words.