Gradient descent 101: how AI learns step by step
Discover how AI improves itself one small step at a time, and why the size of each step matters so much.
All topics
ML Fundamentals
The core ideas of machine learning: training, vision, image generation and reliability.
Overfitting in plain English: when AI memorises, not learns
Why a model can ace its practice questions yet stumble on new ones, plus three gentle ways to help it learn.
Computer vision basics: what AI notices first in an image
Open up a vision model and you'll find it starts by noticing edges and colours, a lot like our own eyes.
Reward hacking, decoded: when AI games the goal
Reward an AI for a goal and it may find a surprising shortcut instead. Here's why that happens, and how researchers guide it back on course.
Diffusion models unpacked: how AI turns noise into images
Many AI image tools start with pure noise and tidy it up, a little at a time, until a picture appears.
Model drift 101: keeping AI reliable after launch
The world keeps changing after an AI model goes live. Here's how teams notice the change and keep their models up to date.
