The AI Glossary: 10 Terms You Need to Know to Understand Artificial Intelligence

Artificial intelligence has slipped into nearly every corner of daily life — drafting emails, screening job applications, recommending what to watch and powering the chatbots that answer customer service calls. But the vocabulary that surrounds the technology can feel like a foreign language, full of jargon borrowed from computer science, statistics and science fiction.

Here are 10 essential terms that make the conversation easier to follow.

1. Artificial Intelligence

The broad umbrella term for computer systems that perform tasks typically associated with human intelligence — recognizing images, understanding language, making predictions or solving problems. It’s less a single technology than a field of study encompassing many approaches.

2. Machine Learning

A subset of AI in which systems improve through exposure to data rather than through explicit, line-by-line instructions. Instead of a programmer writing rules for identifying a cat in a photo, the system is shown many examples and learns the patterns itself.

3. Neural Network

The architecture underpinning most modern AI. Loosely inspired by the human brain, a neural network passes information through layers of interconnected nodes, each adjusting the signal slightly. “Deep learning” simply refers to networks with many such layers.

4. Large Language Model (LLM)

The engine behind chatbots like ChatGPT, Claude and Gemini. An LLM is trained on enormous amounts of text and learns to predict what word is likely to come next. That deceptively simple mechanic, performed at massive scale, produces fluent paragraphs, working code and passable poetry.

5. Training Data

The material — text, images, audio, video — fed into a model so it can learn. The quality, breadth and legality of training data has become one of the most contested issues in the industry, driving lawsuits from authors, artists and news organizations.

6. Parameters

The internal values a model adjusts during training; in effect, the knobs that encode what it has learned. Models are often described by parameter count — billions or trillions — as a rough proxy for size and capability, though bigger does not automatically mean better.

7. Prompt

The instruction a user gives an AI system. “Prompt engineering” — the craft of phrasing requests to get better results — emerged as a genuine skill, and in some cases a job title, as chatbots went mainstream.

8. Hallucination

When a model generates information that sounds authoritative but is false — invented citations, fabricated quotes, imaginary case law. Because language models are built to produce plausible text rather than verified facts, hallucination remains a persistent risk and the main reason AI output should be checked.

9. Generative AI

AI that creates new content — text, images, music, video — rather than only classifying or predicting. It’s the branch responsible for the current boom, and for the debates over deepfakes, copyright and creative labor that have followed.

10. AGI (Artificial General Intelligence)

A hypothetical system that could match or exceed human ability across virtually any intellectual task, rather than excelling at narrow ones. Whether AGI is imminent, decades away or impossible is among the most heated disputes in technology — and it shapes how companies pitch themselves and how regulators weigh the risks.

Why it matters

Understanding these terms isn’t about becoming a programmer. It’s about being able to read a headline, evaluate a product claim or follow a policy debate without taking the marketing at face value. As AI becomes embedded in hiring, medicine, education and the news itself, basic fluency in its vocabulary is quickly becoming a form of literacy. Read More