ai engineer 2026 · part 1 · the map

Every link from the video. Nothing skipped.

The exact things to learn, in order, and where to learn them. Most of this costs nothing. Where something is paid, it says so.

The myth first, because it saves you a year: AI engineers do not build AI. The big models already exist. The job is building things with them.
STEP 1

One language

Pick Python and stop shopping. You do not need the whole language. You need these six things, and then you start building.

CS50P, Harvard free

Harvard's introduction to programming with Python, taught by David Malan. Covers exactly the list above plus libraries, file handling and tests. Free to watch and work through.

cs50.harvard.edu/python ↗

The official Python tutorial free

Straight from the source. Dry, but it is the reference you keep coming back to.

docs.python.org/3/tutorial ↗

freeCodeCamp free

Hands-on lessons in the browser plus long project videos. Good if you learn by typing along rather than watching.

freecodecamp.org ↗

If you pick JavaScript instead free

Also a valid choice, especially if you want your projects to be websites people can click. Same six-item list applies.

javascript.info ↗ theodinproject.com ↗
STEP 2

Drive a model from your code

An API key is just a password that lets your code talk to a model. Get one, send a message, print the answer. That is your first AI app, and it is about ten lines.

Google AI Studio free tier

The fastest free start. Create a key in the browser, no card needed. Free limits are modest and change over time, so check the current ones on the page.

aistudio.google.com ↗

Claude API quickstart paid credits

The official get-started guide. Copy-paste examples in Python, TypeScript and more. Usage is paid, so treat it as the step-up once your free experiments work.

platform.claude.com/docs ↗

Ollama, models on your own machine free

Runs open models on your own computer. No key, no bill, nothing leaves your laptop. Great for practising without watching a meter.

ollama.com ↗

Then three upgrades, in this order:

STEP 3

Ship in public

The step almost everyone skips, and the only one that compounds. One small thing a week. Small really is fine.

GitHub free

Where the code lives. A public repo with a readme explaining what it does and why is already more than most people have.

github.com ↗

Vercel free tier

Where the live link comes from. Free hobby tier is enough to put a working thing on the internet that anyone can open.

vercel.com ↗

Hugging Face free

Models, datasets and free Spaces to host small demos. Useful once you start wanting an audience for the thing you built.

huggingface.co ↗
NEXT

The rest of the series

Part 2 goes deep on step one: which language to actually pick, and the first thing to build with it. New part every week.

Questions about this? Ask in the Discord.

Free tools, honest feedback on your work, and people who answer. Job hunters, builders and learners.

Join free