AI training glossary
128 terms explained in plain English.
A
C
- ChatbotSoftware that holds a text or voice conversation. Older chatbots follow scripted paths; AI chatbots generate their replies with a language model.AI fundamentals
- Context windowThe amount of text an AI model can take into account at one time, including the conversation, attached files and its own replies.AI fundamentals
F
G
H
K
L
M
- Machine learningA branch of AI in which systems learn patterns from examples instead of being programmed with explicit rules. Large language models are one type of machine learning.AI fundamentals
- ModelThe trained AI system that generates the output. The app you type into is the interface; the model is the engine behind it, and one app can offer several models.AI fundamentals
- Multimodal AIAI that can work with more than one kind of input or output, such as text, images, audio and files, in the same conversation.AI fundamentals
P
R
- Reasoning model(Thinking model)A model that works through a problem in steps before giving its final answer. It is slower and uses more tokens, and it tends to do better on multistep tasks.AI fundamentals
- Retrieval-augmented generation(RAG)A setup in which an AI tool first searches a set of documents for relevant passages, then uses them to write its answer.AI fundamentals
T
- TokenA small unit of text, often a word or part of a word, that AI models read and generate. Usage limits, context windows and API pricing are all counted in tokens.AI fundamentals
- Training dataThe text, images and other material used to build an AI model. It shapes what the model knows, how it writes and which biases it may carry.AI fundamentals
Commonly confused terms
AI literacy is understanding what AI is and what it can and can’t do. AI fluency is the practical ability to work with it well on your own tasks. Literacy comes first; fluency comes from practice.
Upskilling improves how someone does their current job. Reskilling prepares them for a different job.
A workshop is hands-on: participants practice on real tasks. A masterclass is expert-led and deep on one topic, with more demonstration than practice. “Training” is the umbrella term for both.
An assistant responds to each request and waits for the next one. An agent takes a goal, plans the steps and uses tools to carry them out with less direction.
An outcome states what the learner can do afterward. An objective states what the trainer aims to achieve. Many organizations use the two interchangeably.
A certificate of completion shows someone finished a course. A certification shows they passed an assessment against a standard. Accreditation is when an external body approves the course or provider itself.
E-learning is delivered entirely through digital platforms. Blended learning combines online study with in-person sessions. Live online training is real-time and trainer-led over video.
A self-paced course runs on each learner’s schedule. A cohort-based course moves a group through together on fixed dates, with live sessions and peer discussion.