The big picture
How to Think About Machine Learning.
AI, ML and LLMs get talked about like magic. They aren't. Underneath, machine learning is a simple idea: instead of writing the rules yourself, you show a computer enough examples that it finds the rules — and everything it works with, images, spreadsheets, language, is just numbers. This is a short, plain-English series that builds the whole mental model from the ground up.
The mental model
The whole idea, in three moves.
Almost everything in the series comes back to these three. Hold them, and the rest — neural networks, LLMs, all of it — is just detail on top.
Learn from examples
You don’t program the rules. You show the computer many examples of a problem and its answers, and it works out the pattern that connects them.
It’s all numbers
Images, text, spreadsheets — inside a computer they’re all just grids of numbers. Once everything is numbers, a model can do maths on it.
Shrink the error
Learning is trial and error made precise: the model measures how wrong it is with a single number, then adjusts itself, over and over, to make that number smaller.
Why it's worth understanding
You don't need the maths to have the right picture in your head.
Most people either treat AI as magic or drown in jargon. There's a middle path: a clear, honest intuition for what these systems really do. That's what lets you use them well, judge the hype, and know what's actually possible.
The series
How to Think About Machine Learning.
Eighteen short episodes across two seasons, each on one idea, in plain English. Season 1 builds the foundations; Season 2 gets to neural networks and LLMs. Open any episode for a one-line summary and the video.
ML Foundations
Parts 1–11 · Season 1Start here. The core intuition behind all of machine learning — what it is, how a model learns, and why underneath it's all just numbers and curve-fitting.
Watch the whole thing in order.
18 episodes · 2 seasons — the full series as one playlist.
A note on the series: this is a simple, high-level take built for intuition, not precision — so some specifics may differ from more formal explanations. That's the point. Get the right picture first; the rigour is easier to add once you have it.