M0: AI Engineering, explained
Before you build, it helps to know what you're building with. "AI engineering" sounds like it needs a PhD and a supercomputer, it doesn't. By the end of this short orientation you'll be able to say, in plain words, what an AI engineer actually does, how a large language model works under the hood, and how to pick the right model for a job, and you'll have compared two models with your own hands. Then we start writing code.
Today's win: you can explain what AI engineering is, roughly how an LLM works, and how to choose a model, and you've run the same prompt on two different models and compared them.
Today you will
- Define AI engineering and how it differs from an ML engineer / data scientist / researcher
- Understand how an LLM works from a builder's view: tokens, context window, training, inference, and why it sometimes makes things up
- See the model landscape (closed vs open, the big families) and how to choose one
- Map how the course's pieces fit: prompting → RAG → agents → fine-tuning
Run of show (~45 min)
| Time | What we do |
|---|---|
| 0:00 | Hook + the win we're chasing |
| 0:05 | The big ideas: what AI engineering is; how LLMs work; the model landscape (full read in notes.md) |
| 0:15 | Lab: "meet the models": run one prompt on two models, compare (browser only) |
| 0:35 | Show: post your comparison + one thing that surprised you |
| 0:40 | Wrap + what's next (M1: your first Python) |
If you get stuck
- This module is reading + a browser activity, no install, no code, no account or key needed. Nothing here can harm your computer.
- There are no "wrong" answers in the lab, you're observing and comparing. Re-read the You should now see line and compare with your breakout partner.
- New term you don't recognize? It's in
resources/glossary.md.
Optional challenge
Find the knowledge cutoff of one model: ask it about a very recent event and see whether it knows, refuses, or makes something up. That "it doesn't know recent or private things" is the exact gap RAG (M7) fixes, you'll have felt the problem before you solve it.