Fundamentals¶
Six short pages that carry the whole year. Every later module is one of these ideas done with more rigour.
Read them alongside the first project, not before it. The words land better once you have seen a model work and then break.
-
What AI is
Where the term came from, what people have meant by it over seventy years, and how AI, machine learning, deep learning and generative AI fit inside each other.
25 min
-
How machines learn
The switch from writing rules to learning them. Every term you need for the first project, and the loop you will repeat all year.
30 min
-
Framing a problem
Turning "can we use AI for this" into something buildable. Five questions, the learning families, and one example where the right answer is no.
25 min
-
Working with data
Splits, variety, sources, and what to check before you build anything. Read this before you collect a single image.
25 min
-
Ethics and responsibility
Consent, bias, and accountability treated as engineering decisions, because that is what they are.
25 min
-
What training actually does
Weights, loss, gradient descent, and the graph that tells you when to stop. Read alongside the maths module.
30 min
These pages are your reference for the year
When a later module uses a term you half remember, come back here rather than searching for it. The definitions on these pages are the ones the rest of the handbook assumes.