What is the difference between AI, machine learning and deep learning?

People use these three terms as if they mean the same thing, but they are actually layers that sit inside each other, like a set of boxes.

Artificial intelligence is the biggest box. It covers any technique that makes a computer behave in a way that looks intelligent. Even an old chess program built entirely from hand-written rules counts as AI. The goal is the behaviour, not a specific method.

Machine learning is a box inside AI. Here, instead of writing the rules, you give the computer data and let it learn the rules itself. A bank that wants to catch fraud can feed past transactions marked "genuine" or "fraud" into a machine learning model. The model notices patterns, such as unusual locations or odd timings, and uses them to judge new transactions. Classic machine learning methods include decision trees, linear regression and random forests. They work well with structured data, the kind you would keep in an Excel sheet.

Deep learning is a smaller box inside machine learning. It uses artificial neural networks with many layers, which is where the word "deep" comes from. Each layer picks up a slightly more complex pattern than the one before. In image recognition, early layers might detect edges, middle layers detect shapes like eyes or wheels, and later layers recognise a whole face or a car. Deep learning needs a lot of data and a lot of computing power, usually GPUs, but it is extremely good at messy, unstructured data like photos, speech and natural language.

Large language models such as the ones behind ChatGPT, Claude and Gemini are deep learning models. So when you chat with one of them, you are using deep learning, which is a type of machine learning, which is a type of AI.

A quick example ties it together. Imagine a system that sorts mangoes by quality. A rule-based AI might say "reject if the weight is under 150 grams." A machine learning model might learn from records of size, weight and colour readings. A deep learning model might look directly at photos of the mangoes and learn to spot bruises on its own.

If you are just starting out, there is no need to master all three at once. Understand the basic idea of learning from data first, and the rest becomes much easier to follow.