AI Foundations
Working with LLMs
The five ideas every AI builder needs
You know what an LLM is. Now you'll learn how to work with one well. These five ideas will come up in every app you build in this course.
Working with LLMs
Think of an LLM as a very fast new assistant. To get great work from them, you need to know a few things:
- How to give clear instructions (prompts)
- How they read and count your words (tokens)
- How much they can keep in mind at once (context window)
- When they might get things wrong (hallucinations)
- Which assistant to pick for which job (model selection)
What you'll learn
Prompts
The instructions you give an AI
Tokens
How AI reads and counts your words
Context window
How much an AI can remember at once
Hallucinations
When AI sounds sure but is wrong
Model selection
Picking the right model for the job
How to use this section
Each idea builds on the one before it, so I recommend going in order. Every page follows the same format:
- What is it? — A simple explanation with an everyday analogy
- See it in action — Example prompts you can copy and try
- Try it yourself — Short exercises to check your understanding
- Common mistakes — Things to watch out for
These ideas are the difference between "AI gave me something random" and "AI built exactly what I wanted". Take your time here. It pays off when you start building.
Practice
Try these exercises to lock in what you just learned.
Quiz
Not startedWhich of these is NOT one of the five ideas in this section?
What's next?
Let's start with the most important skill of all: writing a good prompt.