AI from the first bit: series index

This series starts at the very beginning, with how a computer stores a letter, and from there it climbs toward what happens inside an AI system. Along the way you do not need to install anything or know how to program. Each article is built around a tool that runs in the browser, and you use it before reading the explanation.

The articles are published little by little, and this index grows with them: it lists only what you can already read.

What makes it different

First you try it, then it is explained. Each article begins with a link that opens a tool already set up with the example: you move a slider, change a bit or drag a point, and you see what the article goes on to explain actually happen. The theory comes once the thing it describes is already in front of you.

It all happens in your browser. There is nothing to install, no account to create and nothing to pay. What you type or upload into the tools does not leave your computer. Some of them, the ones that use real models, download a model the first time you use them, and they warn you before doing so and tell you how big it is.

You also learn from what fails. Each article has a section, The classic mistake, with the misunderstanding that comes up most often about that topic and a way to see it in the tool. A model that gets it right teaches little; one that gets a specific case wrong teaches you why.

What it is not

It does not replace learning to program or a PyTorch course. It explains what happens inside an AI system, so that you can judge one; to train large models for real you also need Python and a graphics card. When an article reaches that limit, it will tell you.

How each article is built

• Try this first: the link to the tool, with the example set up, and what to look at.
• What just happened: the explanation.
• The classic mistake: the misunderstanding the article aims to clear up.
• Try it yourself: something to do on your own, with no solution written out, that you can check with the same tool.
• Further reading: where to go deeper, with a note on the level each link requires.

You can read them in order, which is how they were designed, or one at a time: each one makes sense on its own.

The articles

Level 0 — How a machine sees the world

For complete beginners. No math: so that “data”, “model” and “prediction” stop being empty words.

1. 1Everything is a number (yes, even your name)✓ Viewed→
2. 2An image is a table of numbers✓ Viewed→
3. 3A corrupted bit: how to fix it without knowing which one it is✓ Viewed→
4. 4What is a model (and why a straight line already counts as one)✓ Viewed→