Large language model
Also called: LLM
In one sentence
A model that predicts the next token of text, trained on very large amounts of written material.
In more depth
Given the text so far, the model produces a probability distribution over every token it knows. One token is selected, appended to the input, and the process repeats — which is how a whole paragraph gets generated one piece at a time.
Everything else follows from this: fluency, because plausible continuations are exactly what it optimises; confident errors, because nothing in the loop checks truth; and the leverage of prompting, because the preceding text is the only lever on what comes next.