AI words, without the jargon
Short, plain definitions we keep up to date. Each one links to the articles that use it.
- AI Act (EU)The European Union's law on artificial intelligence. It sets rules for AI systems based on how much risk they pose to people's safety and rights.
- AI agentAn AI system that can take actions to complete a task, not just answer questions. It may search the web, use apps or carry out several steps on its own.
- AlgorithmA set of step-by-step instructions for solving a problem or completing a task. Computers follow algorithms to do almost everything they do.
- AlignmentThe work of making sure AI systems do what people actually intend and act in line with human values, rather than causing harm or pursuing the wrong goal.
- Bias (in AI)When an AI system gives results that are unfair or skewed, for example working better for some groups of people than others, often because of patterns in its training data.
- ChatbotA program you talk to by typing or speaking, which answers in ordinary language. Many modern chatbots run on a large language model.
- Context windowThe amount of text an AI model can take into account at one time, including your messages, any documents you share and its own replies.
- DeepfakeA fake video, image or audio recording made with AI to show a real person saying or doing something they never said or did.
- Fine-tuningExtra training that takes an existing AI model and adapts it to a particular job, style or subject using a smaller set of examples.
- Generative AIAI that creates new content, such as text, images, music, video or computer code, based on patterns it learned from existing examples.
- HallucinationWhen an AI tool states something false or made up as if it were true, such as a fake quote, a wrong date or a book that does not exist.
- Large language model (LLM)An AI model trained on huge amounts of text so it can read and write language. It works by predicting which words are likely to come next.
- Machine learningA way of building AI in which a computer learns patterns from examples, instead of a programmer writing every rule by hand.
- Model (AI model)The trained part of an AI system that turns an input, such as a question or a photo, into an output, such as an answer or a label.
- MultimodalDescribes an AI system that can work with more than one kind of input or output, such as text, images, sound and video.
- Neural networkA type of machine learning system made of many simple connected units that pass numbers to each other. It is loosely inspired by how brain cells connect.
- Open-weights modelAn AI model whose trained numbers, called weights, are published so anyone can download it and run it on their own computer or server.
- PromptThe instruction or question you give an AI tool, usually typed in ordinary language. It tells the system what you want it to do.
- TokenA small chunk of text, often a short word or part of a word, that a language model reads and writes. Usage limits and prices are often counted in tokens.
- Training dataThe examples an AI system learns from, such as text, images, recordings or records. What is in the data shapes what the system can do and the mistakes it makes.