AI glossary
The words you will meet most often, without the jargon.
- AI (artificial intelligence)
- Computer systems that perform tasks usually associated with human intelligence, such as understanding language or recognising images.
- Machine learning
- A way of building AI where systems learn patterns from data instead of following fixed rules.
- Deep learning
- Machine learning that uses many-layered neural networks to find complex patterns.
- Neural network
- A computing model loosely inspired by the brain, made of connected layers of simple units.
- Large language model (LLM)
- An AI model trained on huge amounts of text to understand and generate language.
- Generative AI
- AI that creates new content such as text, images, audio or code.
- Prompt
- The instruction or question you give an AI tool.
- Token
- A small chunk of text, such as a word or part of a word, that a model reads and produces.
- Context window
- The amount of text a model can consider at one time.
- Hallucination
- A confident but false or invented answer produced by an AI model.
- Training data
- The examples a model learns from.
- Fine-tuning
- Further training of an existing model on specific data for a particular task.
- RAG (retrieval-augmented generation)
- A method where a model looks up relevant documents before answering, to ground its response.
- AI agent
- An AI system that can plan and take actions, such as using tools, to complete a goal.
- Multimodal
- Able to work with more than one type of input, such as text, images and audio.
- Bias
- Systematic unfairness in outputs, often reflecting patterns in the training data.
- Inference
- The stage when a trained model produces an answer to your request.
- Open-weight / open-source model
- A model whose weights, and sometimes code, are published for others to run and adapt, under specific licences.
- Parameter
- An internal value a model adjusts during training; models can have billions of them.
- Alignment
- Work to make AI systems behave in line with human goals and values.