AI Terminology Glossary
Guide to Understanding Artificial Intelligence and AI Terminology
Artificial Intelligence (AI) is a field filled with technical terms and concepts that can be confusing. This AI Terminology glossary explains the most common AI terms in simple, easy-to-understand language. Whether you’re a student, a small business owner, or just curious about AI, this guide will help you understand the basics and beyond.
A
Algorithm
A step-by-step process or set of rules that a computer follows to solve a problem or make a decision.
Example: An algorithm might help a chatbot decide how to answer your question.
Artificial General Intelligence (AGI)
A type of AI that could think and learn like a human, capable of solving many different kinds of problems. This is still a future concept.
Artificial Intelligence (AI)
The ability of a computer or machine to perform tasks that normally require human intelligence, such as understanding language, recognising images, or solving problems.
Artificial Neural Network (ANN)
A system of algorithms inspired by the human brain, designed to process data in complex ways, like recognising patterns in images.
B
Bias (AI)
When an AI system makes unfair or inaccurate decisions because of problems in the data it was trained on.
Example: AI that performs worse for certain groups of people because of unbalanced training data.
Big Data
Large and complex datasets that AI systems analyse to find patterns and make decisions.
Bot
A program designed to perform tasks automatically. In AI, bots often interact with humans, like customer service chatbots.
C
Chatbot
An AI-powered program that simulates human conversation. It can answer questions, provide support, or help you make decisions.
Classification
A type of machine learning where AI sorts data into categories.
Example: Sorting emails into “Inbox” and “Spam.”
Computer Vision
A field of AI that helps computers understand and interpret visual information, like recognising objects in images.
Context Window
The amount of text or information an AI model can process at one time when generating a response.
Conversational AI
AI designed to hold natural conversations with people, such as virtual assistants like Siri or Alexa.
D
Dataset
A collection of data that AI uses for training or analysis.
Example: A dataset might include thousands of pictures of dogs to help an AI recognise different breeds.
Deep Learning
A type of machine learning that uses neural networks to learn from large amounts of data. It’s often used for tasks like image recognition and language translation.
Diffusion Model
An AI system that generates images by reversing noise added to an initial image, often used in tools that create art from text descriptions.
E
Edge AI
AI systems that process data locally on devices like smartphones, instead of relying on cloud servers. This makes tasks faster and more private.
Ethics in AI
The study of how to use AI responsibly and avoid harm, focusing on issues like fairness, privacy, and transparency.
Expert System
A computer program designed to solve complex problems by mimicking human decision-making.
F
Fine-Tuning
The process of training a pre-built AI model on a specific type of data to make it work better for a particular task.
Foundation Model
A broad AI model trained on massive datasets that can be adapted for many tasks.
Example: OpenAI’s GPT is a foundation model used for chatbots and other applications.
Frontier Model
An advanced AI model that represents the latest innovations and capabilities in the field, often still in development.
G
Generative AI
AI that creates new content, such as text, images, or music, based on patterns it learned from training data.
Example: AI generating a painting based on your description of a landscape.
GPU (Graphics Processing Unit)
A type of computer chip that helps AI process data quickly, especially for tasks like training large models.
H
Hallucination (AI)
When an AI confidently generates incorrect or nonsensical information because of gaps in its training.
Example: A chatbot making up facts in response to a question.
Hyperparameters
Settings that guide how an AI model is trained, like how much data to process at once or how fast it should learn.
I
Image Recognition
The ability of AI to identify and classify objects, people, or scenes in images.
Inference
The process of using a trained AI model to make predictions or generate results based on new input data.
Insightful AI
A business dedicated to providing AI-powered solutions for small businesses, focusing on accessibility and practical implementation.
Intent Recognition
AI’s ability to understand the purpose or goal behind a user’s input, often used in chatbots and virtual assistants.
L
Large Language Model (LLM)
An AI model trained on vast amounts of text to understand and generate natural language, like GPT or Google’s Gemini.
Learning Rate
A setting that determines how quickly an AI model adjusts its parameters during training.
Logistic Regression
A simple machine learning method used for classification tasks, like predicting whether an email is spam.
M
Machine Learning (ML)
An AI model trained on vast amounts of text to understand and generate natural language, like GPT or Google’s Gemini.
Model
The trained system that performs a specific AI task, like generating text or recognising images.
Multimodal AI
AI that can process and understand different types of data, such as text, images, and video, all at the same time.
N
Natural Language Processing (NLP)
A field of AI that enables computers to understand and respond in human language.
Example: Translating text or answering questions in chatbots.
Neural Network
A system of algorithms inspired by the structure of the human brain, designed to process and learn from data.
P
Parameters
Internal values that an AI model adjusts during training to improve its accuracy.
Personalisation
The use of AI to tailor experiences, like recommending products or customising marketing messages, based on individual preferences.
Predictive Analytics
Using AI to analyse historical data and predict future trends, like sales forecasts or customer behaviour.
R
RAG (Retrieval-Augmented Generation)
A method where AI retrieves external data to improve its answers, reducing mistakes or “hallucinations.”
Reinforcement Learning
A training method where AI learns by receiving rewards or penalties for its actions, helping it improve over time.
Regression
A type of machine learning used to predict continuous outcomes, like the future price of a product.
S
Scalability
The ability of an AI system to handle larger amounts of data or more users without losing efficiency.
Sentiment Analysis
AI’s ability to determine the emotion or opinion behind text, like identifying whether a customer review is positive or negative.
Supervised Learning
A type of machine learning where AI is trained on labelled data, meaning the correct answers are provided during training.
T
Token
Small chunks of data, like words or parts of words, that AI analyses to understand and generate text.
Transformer
A type of neural network architecture used in many AI models, enabling them to process sequences of data like text.
Training Data
The information AI uses to learn patterns and improve its performance.
U
Unsupervised Learning
A type of machine learning where AI is trained on unlabelled data, allowing it to find patterns or groupings on its own.
Upscaling
Using AI to enhance the quality of images or videos, making them clearer and more detailed.
V
Virtual Assistant
An AI-powered program that performs tasks or provides services based on voice or text commands.
Example: Siri, Alexa, or Google Assistant.
Voice Recognition
AI’s ability to recognise and process spoken language, often used in virtual assistants or transcription tools.