A word at a time

Glossary

You do not need to memorise these. Look up a word when it gets in your way, then return to what you were trying to do.

AI: Computer systems that perform tasks such as recognising patterns or working with language. A spam filter and a writing assistant can both use AI. See it in practice →
Generative AI: AI that produces content such as text, images or audio from input. A new paragraph is not necessarily a verified fact. See it in practice →
Large language model: A model trained on large amounts of language to generate responses. It can produce useful text and also confident errors. See it in practice →
Prompt: The request or instruction you give an AI tool, such as “summarise these notes in three bullets”. See it in practice →
Context: Relevant information that helps interpret a request: who the message is for, confirmed facts and what happened earlier. See it in practice →
Hallucination: An invented or unsupported AI response presented as if it were true, such as a citation that does not exist. See it in practice →
AI assistant: A product you interact with to ask questions or complete tasks. Its underlying models and tools may change. See it in practice →
Agent: An AI system configured to take steps towards a goal, sometimes using tools. Its permissions, stop conditions and human oversight matter. See it in practice →
Workflow: An ordered process for getting something done, such as receiving a request, recording it and replying. It can be manual or automated. See it in practice →
Automation: Software carrying out repeatable steps when a trigger occurs. It does not necessarily use AI. See it in practice →
Trigger: The event that starts a workflow, such as a new form submission. See it in practice →
API: A defined way for software to request data or actions from another system. A connector may use an API on your behalf. See it in practice →
Webhook: A message one system sends to another when an event happens, such as a submitted form. See it in practice →
Vibe coding: Building software by describing what you want to an AI tool and iterating on its code. Testing and judgement still matter. See it in practice →
Frontend: The part of a website or app that people see and interact with in their browser. See it in practice →
Backend: Server-side software that handles work such as permissions or processing requests. Hiding a button in the frontend is not backend security. See it in practice →
Database: An organised store of information that software can read and update. A browser-only calculator need not have one. See it in practice →
Domain: A name used to find a website, such as example.com. A domain name is separate from the site files and hosting. See it in practice →
Hosting: The service or infrastructure that serves a website or application to visitors. See it in practice →
Deployment: Putting a particular version of a website or application onto its hosting environment. A local preview is not a public deployment. See it in practice →
Reference image: An image used to guide a subject, composition or visual style. State what it should guide and check permission before uploading it. See it in practice →
Text-to-image: Creating an image from a written description. It may take several revisions to match the brief. See it in practice →
Image-to-video: Creating motion from an input image. A reference helps establish a scene but does not guarantee consistency. See it in practice →
Text-to-video: Creating a video clip from a written description. Available duration and control vary by tool. See it in practice →
Aspect ratio: The relationship between image width and height. For example, 9:16 is tall and 16:9 is wide. See it in practice →
Iteration: Making a version, checking it and improving a specific part. Change one thing when you want to understand what helped. See it in practice →

Try explaining one term to a friend using your own example. If it still feels abstract, follow its lesson link and see it in context.

Back to learning →Find a useful template