A Family AI Glossary: Twenty Words Explained in One Sentence Each

The basic building blocks

These first five terms provide a foundation for conversations about artificial intelligence. They describe the technology itself and the information it works with. A useful way to explain AI to parents is to connect each word to something already familiar, such as a recipe, a conversation or a filing cabinet.

The definitions are deliberately short. The examples help show where the terms appear in ordinary digital life.

• 1. Artificial intelligence (AI): Computer technology designed to perform tasks that usually involve abilities such as recognising patterns, understanding language or making predictions.
• 2. Machine learning: A way for computers to identify patterns from examples and use those patterns when handling new information.
• 3. Model: The computer system that has been trained to recognise patterns and perform a particular AI task.
• 4. Training data: The examples and information used to develop an AI model.
• 5. Algorithm: A set of instructions or mathematical steps a computer follows to complete a task.

Words used when creating content

Generative AI is responsible for many of the tools families now encounter in writing, image creation and entertainment. These terms describe how people communicate with such tools and the kinds of material they can produce.

The distinction between generating and retrieving information is especially useful. A system that creates a new paragraph is doing something different from a service that finds an existing article.

• 6. Generative AI: Artificial intelligence that creates new material, such as text, pictures, music or video.
• 7. Prompt: A question, instruction or description given to an AI tool to guide what it produces.
• 8. Chatbot: A computer programme that communicates with people through a conversation-style interface.
• 9. Large language model (LLM): An AI model trained to recognise patterns in language and produce text in response to instructions.
• 10. AI-generated content: Text, images, audio or other material created wholly or partly with an AI system.

Words about how AI handles information

Some AI vocabulary describes what happens behind the conversation. Understanding these terms helps explain why an answer may sound confident while still containing an error. It also gives families a better way to discuss how much information a tool receives and uses.

For example, an AI assistant may use information from the current conversation to shape its next response. That does not automatically mean the information is permanently remembered or used in every future conversation.

• 11. Data: Information that a computer can store, process or use.
• 12. Dataset: An organised collection of information used for analysis or AI development.
• 13. Context: The surrounding information that helps an AI system interpret a question or instruction.
• 14. Hallucination: An AI-generated statement that sounds plausible but contains invented or incorrect information.
• 15. Personalisation: Adjusting a digital service or its responses to reflect an individual's preferences or information.

Words about trust, safety and responsibility

The final five terms are particularly useful when relatives discuss AI-generated photographs, suspicious messages or online services. They cover questions about identity, information handling and the decisions people make when using technology.

These concepts should be considered separately. A photograph can be AI-generated without being fraudulent, while an ordinary photograph can be used in a misleading message. Understanding the distinction helps families ask better questions.

• 16. Deepfake: An image, video or audio recording manipulated or generated with AI to imitate a real person or event.
• 17. Synthetic media: Images, recordings or videos created or substantially altered using computer-generated techniques.
• 18. Privacy: The ability to control how personal information is collected, used, shared and stored.
• 19. Bias: A systematic tendency in information or a computer system that can produce unfair or unbalanced outcomes.
• 20. Human oversight: A person remaining involved in decisions, checking information or supervising how an AI system is used.

How to use the glossary in a family conversation

A glossary is most useful when people use it to clarify a real question. Choose three unfamiliar words from the list and ask each relative to explain one in their own words. Different explanations can reveal which part of a concept needs another example.

For instance, someone might describe a prompt as a question asked of a computer, while another person might call it an instruction. Both descriptions can be useful because prompts can take either form. A request such as 'Explain how to grow tomatoes in five steps' is a prompt, even though it is not phrased as a question.

The glossary can also help when a family member encounters a new AI feature. Instead of treating every unfamiliar term as a separate mystery, identify whether it concerns the model, the information provided, the content created or the way the service handles personal details.

For a simple exercise, write five terms on separate pieces of paper and place them beside a shared device. Ask one person to choose a term and another to give an everyday example. Finish by choosing one AI feature to explore together and identifying the words that describe it.

Questions readers ask