AI, machine learning and deep learning - how they nest
What is artificial intelligence?
AI is the broad field of building machines that perform tasks we'd normally consider to need human intelligence - understanding language, recognising images, making decisions, playing games. It's a goal, not a single technique.
What is machine learning, and how is it different from AI?
Machine learning is a subset of AI - the dominant approach today - where a system learns patterns from data rather than being explicitly programmed with rules. Instead of a human writing 'if this, then that' for every case, you show the model many examples and it infers the rule.
What is deep learning?
Deep learning is a subset of machine learning that uses neural networks with many layers. It's especially good at messy, high-dimensional data - images, audio, natural language - and powers most of the recent breakthroughs, including modern chatbots.
So how do the three relate?
They nest like Russian dolls: deep learning is inside machine learning, which is inside AI. All deep learning is machine learning; all machine learning is AI; but not all AI is machine learning (early AI used hand-written rules).
How machine learning actually learns
The key shift machine learning made is this: instead of a programmer writing every rule, the system learns the rules from examples. Show it thousands of labelled photos of cats and dogs and it learns the patterns that distinguish them - no one hand-codes 'a cat has pointy ears'.
There are three broad styles. In supervised learning the data is labelled and the model learns to predict the label (spam or not spam). In unsupervised learning the data is unlabelled and the model finds structure (grouping similar customers). In reinforcement learning an agent learns by trial and error from rewards (mastering a game). More data and better data usually beat cleverer algorithms.
Where you already use AI
AI stopped being futuristic the moment it became invisible. You almost certainly used several examples today without noticing.
- ✓Recommendations: what streaming services, shops and social feeds suggest next.
- ✓Voice assistants and speech-to-text: understanding and transcribing what you say.
- ✓Spam filters, fraud detection and search ranking: quietly sorting and protecting.
- ✓Maps and navigation: predicting traffic and the fastest route.
- ✓Chatbots and writing assistants: answering questions and drafting text - like generating a quiz on any topic.
What generative AI changed
Most AI of the past decade was predictive - it classified or forecast (is this spam? what's the traffic?). Generative AI flipped that: it creates new content. Large language models (LLMs) trained on vast text can write, summarise, translate, answer questions and code; other models generate images, audio and video.
This is what put AI directly in ordinary hands. You no longer need to be an engineer to use it - you type a request in plain language and get a useful result. That's exactly how an AI quiz generator works: you describe a topic, and a model writes tailored questions with explanations in seconds.
Practice this now