What Is Artificial Intelligence?

"Artificial Intelligence" gets used to describe everything from a chess program to a chatbot to a self-driving car, which makes it feel like a buzzword. A useful working definition: AI is the field of building computer systems that perform tasks which normally require human intelligence — things like recognizing images, understanding language, making decisions, or learning from experience.

Narrow AI vs. general AI

Almost every AI system you interact with today is narrow AI (also called "weak AI"): it is very good at one thing — translating text, recommending videos, detecting spam, generating an image — and useless outside that lane. A spam filter cannot drive a car, and a chess engine cannot write an email.

General AI (or "AGI", artificial general intelligence) refers to a hypothetical system that could learn and reason across any domain the way a human can, transferring knowledge from one kind of problem to a completely different one. It does not exist yet, and there's no consensus on when — or whether — it will. Keep this distinction in mind: it's why a large language model can write impressively about a topic yet still fail at simple logic puzzles it wasn't shaped to handle.

AI, machine learning, and deep learning

These three terms are often used interchangeably, but they are nested, not synonymous:

• Artificial Intelligence — the broad goal described above, including old-fashioned approaches like hand-coded "if this, then that" expert systems.
• Machine Learning (ML) — a way of achieving AI where, instead of writing explicit rules, you let the computer find patterns in data. This is where most of modern AI actually lives, and it's the subject of the next several modules.
• Deep Learning — a family of machine learning techniques based on neural networks with many layers ("deep" networks). Deep learning is why AI got dramatically better at vision, speech, and language over the last decade.

A very short history

AI has gone through cycles of hype and disappointment (sometimes called "AI winters") since the 1950s. The two things that changed the trajectory most recently were more data (the internet) and more compute (GPUs), which let deep learning techniques — some of them decades old — finally work well in practice.

Why this matters for the rest of this guide

Because modern AI is overwhelmingly built from machine learning, the rest of this guide focuses there: how a computer can learn patterns from examples (Modules 3–4), the math that makes it possible (Module 2), specific techniques like linear regression, k-Nearest Neighbors, and neural networks (Modules 5–8), and two of AI's biggest application areas — computer vision and natural language processing (Modules 9–10).