What is LLM?

A Large Language Model (LLM) is a type of AI trained on lots of text to learn patterns in language. LLM stands for Large Language Model: “large” means it has many internal settings, “language” means it works with words, and “model” means a system that makes predictions. It generates text by predicting what words are likely to come next. For everyday users, LLMs power chatbots, writing help, and question answering. A downside is they can sound confident while being wrong, and they may reflect bias from the data they learned from.

It is a computer program that has read a huge amount of writing and learned how words usually go together. When you ask it something, it tries to write a helpful answer by guessing what comes next. It can be great for ideas and explanations, but we still need to check important facts.

Imagine a phone's autocomplete has seen millions of books and websites. When you start a sentence to send a message, it tries to finish it in a way that sounds right. Most of the time it is useful, but sometimes it fills in a sentence with something that only sounds right, not something that is true.