Generative AI Course

Generative AI models don’t just analyze data — they produce brand-new text, images, and code from a prompt. This course covers how large language models and image generators actually work, how to prompt them well, and where they reliably fall short.

What you'll learn

This course is specifically about generative AI — the kind that produces new text, images, or code, rather than the kind that classifies or predicts on data you already have. You'll build real intuition for how large language models turn a prompt into a response one token at a time, how image models turn random noise into a picture, and how the handful of parameters every API exposes (temperature, max tokens, top-p) actually change what comes out.

The second half is deliberately practical and honest: what a real API call looks like, what these models are structurally bad at (and why that isn't a bug waiting to be patched), and how to use generated content responsibly — attribution, disclosure, and treating output as a draft rather than a fact. This course assumes no machine learning background and doesn't require any prior AI course on this site, though basic comfort reading Python helps for the one code-heavy lesson.

Lessons

  1. What Is Generative AI?
  2. How Large Language Models Work
  3. Prompt Engineering Basics
  4. Using an LLM API
  5. Text Generation Parameters
  6. Image Generation Basics
  7. Limitations & Hallucinations
  8. Responsible Use of Generative AI
Where to start: begin with What Is Generative AI? and go in order — later lessons (prompting, parameters, limitations) assume you already have the token-by-token mental model built in lesson 2.