Skip to content

Module 01: AI and LLM Fundamentals

Level: Beginner · Time: ~4 hours including exercises · Exam relevance: CCAO-F D2, D3, D6 · CCAR-P D2, D5

This module assumes you have never used an AI model. By the end you will understand what is actually happening when you type a question into a chat box and a large language model writes back, and why it sometimes gets things confidently wrong. Every later module builds on the vocabulary introduced here: model, training, token, context window, hallucination.

You do not need to write code for this module. A few pages show small Python snippets, but they are optional.

Learning objectives

After completing this module you will be able to:

  1. Explain in plain language what artificial intelligence (AI), machine learning and deep learning are, and how they relate to each other.
  2. Describe what makes generative AI different from AI that classifies or predicts.
  3. Explain how a large language model (LLM) produces text by predicting one token at a time, and what pre-training and fine-tuning contribute.
  4. Define a token, estimate token counts for ordinary text, and explain why tokens matter for cost, speed and limits.
  5. Explain what a context window is, what counts towards it, and what happens when a conversation grows too long.
  6. Define hallucination, give three common causes, and apply at least four techniques to reduce or catch it.
  7. Give a balanced account of what LLMs do well and badly, so you can decide when to use one and when not to.

Page order

# Page What you learn Time
1 What is AI? AI, machine learning, deep learning, and where LLMs fit 25 min
2 Generative AI Models that create new content rather than label existing content 25 min
3 Large language models Next-token prediction, transformers, pre-training and fine-tuning 35 min
4 Tokens and tokenisation The units models read and write, and why they drive cost 30 min
5 Context windows The model's working memory and how to stay within it 30 min
6 Hallucinations Why models invent facts and how to catch it 35 min
7 Capabilities and limitations A realistic picture of strengths and weaknesses 30 min

Read the pages in order. Each ends with a short "Check your understanding" quiz; if you miss more than one answer, reread the "How it works" section before moving on.

Prerequisites

  • None. If you can use a web browser and read English, you are ready.
  • Optional: complete Environment setup if you want to run the small Python examples.

How this module connects to the exams

Both supported certifications assume you understand these fundamentals, even though no exam domain is named "AI basics". The exam facts themselves (item counts, timing, weights) are recorded in registry.yaml.

Page CCAO-F objectives CCAR-P objectives
03 Large language models CCAO-F-3.2 CCAR-P-2.1
04 Tokens CCAO-F-3.3, CCAO-F-3.4 CCAR-P-2.4, CCAR-P-4.5
05 Context windows CCAO-F-3.4 CCAR-P-2.4, CCAR-P-3.8
06 Hallucinations CCAO-F-2.1, CCAO-F-2.2, CCAO-F-2.3, CCAO-F-2.4 CCAR-P-4.4, CCAR-P-5.2
07 Capabilities and limitations CCAO-F-4.5, CCAO-F-6.1 CCAR-P-5.2, CCAR-P-6.2

CCAO-F (Claude Certified Associate – Foundations) is for business and knowledge workers. CCAR-P (Claude Certified Architect – Professional) is for solution architects and senior engineers. See Choose your path if you are unsure which applies to you.

Practice

Next module

Continue to Module 02: Claude Fundamentals, where you apply these ideas to Claude specifically and have your first conversation.