5 min · +70 XP
What an AI model actually does
Prediction, tools, and what the model is not.
By the end of this lesson
You will be able to explain in two sentences what a language model does and how it uses tools.
The AI behind Claude is a large language model, often shortened to LLM. It was trained on a very large amount of text and learned to predict what text should come next.
Your phone suggests the next word from the last few words you typed. A language model does the same kind of guessing, but it learned from far more text and can continue a whole explanation, a plan or a program.
Prediction sounds too simple to be useful. At this scale it is enough to explain, translate, plan and write working code. The model does not look answers up in a database; it produces them each time.
Tools let it act
On its own, a model can only produce text. A program around the model can give it tools: read a file, edit it, run a command, search the web. The model decides which tool to use, the program runs it and passes the result back to the model.
Claude Code is such a program. The official documentation describes it as the layer around the model that provides the tools and manages what the model sees.
It is not a person and not a search engine. It is not always right, and it can sound sure while being wrong. During a conversation the model itself does not change: only the text in front of it does.
At its core, what is a large language model doing?
How does Claude Code manage to change a file on your computer?
Facts checked against the official Claude Code docs October 24, 2026 · source
Author
Evgenii Arsentev
PhD · Chief Executive Officer, digital health