AI, ML, Deep Learning, GenAI — the Map of Terms

In her first week, Iris hears four different words used for what seems like the same thing. The exec deck says "our AI strategy". The data team says "the ML models". A job posting mentions "deep learning experience". Her news feed is all "GenAI". Are these four technologies? Four teams? Four budgets? Nobody around her stops to define anything, because everyone assumes everyone else knows.

Here is the secret: the four words are not rivals or synonyms. They nest inside one another, like boxes inside boxes, and once you see the nesting, every buzzword you will ever meet lands neatly in its place. This page is the map for the whole book — five minutes here saves confusion everywhere else.

AI — the Ambition

Artificial intelligence is the biggest box, and the vaguest on purpose. It names an ambition, not a technology: making computers do things that seem to need human intelligence — recognizing a face, holding a conversation, planning a route. It is a field of study, the way "medicine" is a field. Nobody installs "some AI" on a server, just as nobody prescribes "some medicine"; there is always a specific thing inside the box doing the work.

Over the decades, researchers tried many roads toward that ambition — including, for a long time, writing ever-cleverer hand-made rules. Keep that in mind whenever you read that some system "uses AI": the phrase tells you the goal, and nothing at all about the machinery.

ML — the Method That Won

Inside the AI box sits machine learning — the approach from the previous topic, where the computer finds its own rules from examples. Out of all the roads toward the AI ambition, this is the one that pulled ahead, so decisively that today when a product says "AI", the machinery underneath is almost always ML.

This is why the course you are reading is called Machine Learning from Zero and not AI from Zero. Learning the method that actually powers things beats learning the umbrella word. When Plateful's exec deck says "AI strategy", what the company concretely has is a handful of ML models — the delivery predictor among them.

Deep Learning — ML Built from Neural Networks

Inside the ML box sits a smaller one. Deep learning is machine learning built from a particular kind of machinery called a neural network — layers of simple calculations stacked on top of each other. For now, that one sentence is all you need; Chapter 8 opens this box properly and shows there is no magic in it.

What earns deep learning its own box is what it unlocked. Classical ML — the kind in Chapters 3 through 7 — is excellent with facts in tables: distances, prices, order counts. Deep learning is what finally let machines learn from photos, sound, and language. When your phone recognizes your face, that is this box.

Generative AI — the Newest, Loudest Box

And inside deep learning sits the smallest, newest box, the one making all the noise. Generative AI is deep learning turned to producing things — models that write text, draw images, or compose audio, rather than just labeling or predicting. The chatbot on your phone lives here. So does every "AI-generated" picture in your feed.

Notice what this means: the chatbot is not a break from everything before it. It is the same learning-from-examples idea, scaled up inside box after box. Chapter 9 takes you there step by step — and by then, none of it will look like magic.

Reading the Map

The nesting gives you three quick decoding rules. Everything in a smaller box is also in the bigger ones: every deep-learning system is ML, and every ML system is AI — so a company calling its spam filter "AI" is not lying, just gesturing at the biggest box. But it does not run backwards: plenty of ML is not deep learning, and most AI headlines are really about one small corner, GenAI.

And the older boxes did not empty when the new ones appeared. Classical ML still quietly runs most business predictions — prices, risk, recommendations, Plateful's delivery times. The loudest box is not the largest, and this book spends its time accordingly: Chapters 2 through 7 in the ML box, Chapter 8 stepping into deep learning, Chapter 9 into GenAI.