Learn by doing.
There is far more written about AI than anyone can read, and most of it is selling something. This is the ordered version: a route for whatever you are actually trying to do, and nothing on it that we have not read ourselves.
Free, no sign-up, no gate. Pick the line below that sounds like you.
Start with a path
Each one is three or four things to read and do, in an order that works. An hour gets you through most of them.
- I'm starting from zero Four things to read and do, in order, that take you from never having tried this to making something real.
- I write code Getting real speed out of these tools without ending up with a codebase you cannot defend in review.
- I build product Enough to tell a real use case from a demo, spec it properly, and not be the last person in the room to understand the constraint.
- I run a business Where this genuinely earns its keep in a small company, where it does not, and how to tell which is which before you spend anything.
- I keep hearing about agents What the word actually means, which patterns hold up under load, and why most agent demos never make it to production.
- I want to know how it works Tokens, training, context windows, and why a model states something false with total confidence. No maths required to start.
- I want it to answer from my own documents Getting a model to answer from your documents instead of its training. What it takes, and why the naive version disappoints.
- I want to automate my work Moving from one-off chats to work that happens on its own: spotting the task, wiring it up, and keeping it running.
- I'm building with Claude Past the chat window: projects, skills, tools, and the Agent SDK, plus the patterns worth copying.
- I want to ship software with AI Fast and loose to start, structured once it matters. The workflow for shipping software you would still put your name on.
- I want to do this professionally The systems thinking behind AI that runs in production: inference mechanics, retrieval, evals, observability, and the ways it fails at 3am.
- I want the real foundations The long road: intuition first, then the maths, then building it yourself. Slower than the other paths, and it does not expire.
Or read a guide on its own
Long reads, written once and kept current. The paths above are mostly made of these.
- AI Engineer Fundamentals What actually separates someone who ships AI products from someone who just prompts. The systems thinking behind production-grade AI.
- Git Fundamentals What you need to know to stop breaking things
- How LLMs Actually Work Tokens, training, context, and why they make things up. In plain English, no math required.
- Master AI Without Coding: The Complete Guide The fundamentals you need to know to start your journey on Artificial Intelligence.
- Spec-Driven Development: Code With AI Without Losing Control Write the spec, let AI build it. The workflow that keeps you in charge when AI writes most of the code.
- Vibe Coding: From Zero to Hero Master the art of AI-powered development and transform your coding workflow from traditional to revolutionary.
Reference
What we keep open in a tab while building, sorted by topic. Ours and everyone else's.
The shelf we usePlayground
Image, video, audio and text models in one place, for when reading has stopped being the useful part.
Go and play