This is where I go back to the fundamentals of AI and rebuild them from the ground up, one topic at a time, explained the way I came to understand them.
It is not a book with chapters to read in order. The map below covers the whole territory, from classical search algorithms to the internals of large language models, and I explore it in whatever order curiosity takes me. Each topic gets its own standalone note, linked to the ideas it builds on.
- Explained my way. Text, drawings, diagrams, interactive demos or code written from scratch, whatever made the idea click for me.
- Built to be understood. Where it helps, I implement things from scratch before reaching for a framework.
- Growing over time. Notes are marked 🌱 seed, 🌿 growing or 🌳 solid, and get revised as later topics teach me something new.
- Open questions included. Every note ends with what I still don't understand. That is part of the process.
The map
Click a node to expand or collapse it, scroll to zoom and drag to move around. Topics with a published note become links.