Artificial intelligence comes with a lot of jargon. This glossary explains common AI terms in plain language, as of October 2026. Use it as a quick reference while reading about AI, and follow the links to our detailed guides.
Basics
Artificial intelligence (AI): technology that lets computers perform tasks that normally need human intelligence. See our guide to AI.
Super Intelligence (SI): the term the US government adopted in September 2026 for what is commonly called AI in its official communication. It refers to the same technology.
Machine learning (ML): the main way modern AI works: computers learning patterns from data rather than following only fixed rules.
Deep learning: machine learning using neural networks with many layers, behind most recent AI advances.
Neural network: a computing system loosely inspired by the brain, made of connected layers of simple units.
Algorithm: a set of step-by-step instructions a computer follows.
Models and how they work
Model: the trained AI system that makes predictions or generates content.
Large language model (LLM): a model trained on huge amounts of text that can write, summarise and answer questions.
Transformer: the model architecture, introduced in 2017, behind most modern language models.
Parameters: the internal settings a model adjusts during training; large models have billions.
Training: teaching a model by processing data. Inference: using a trained model to answer requests.
Fine-tuning: further training a model on specific data for a particular task.
Token: a small piece of text, such as a word or part of a word, that a language model processes.
Context window: how much text a model can consider at once.
Multimodal: able to work with several types of data, such as text, images and audio.
Open-source model: a model whose weights are publicly released for others to use.
Using AI
Prompt: the instruction or question given to an AI system.
Generative AI: AI that creates new content. See our guide.
Chatbot: a program that converses with people. See What is a bot.
AI agent / agentic AI: AI that plans and takes actions to achieve goals. See our guide.
RAG (retrieval-augmented generation): giving an AI relevant documents to read before it answers, improving accuracy.
Hallucination: a confident but incorrect answer from an AI.
Prompt injection: hidden instructions that try to trick an AI into unintended actions.
Risks and governance
Bias: unfair patterns AI learns from data.
Deepfake: AI-generated media imitating a real person.
Alignment: making AI systems act according to human intentions and values.
AGI (artificial general intelligence): hypothetical AI matching humans across most tasks; it does not exist yet and is debated.
EU AI Act: the European Union's risk-based AI law. See Who controls AI.
Infrastructure
GPU: a chip that performs many calculations in parallel, central to AI. See AI chips.
NPU: an AI processor built into phones and laptops.
HBM: high-bandwidth memory stacked close to AI chips.
Data center: a facility housing servers, power and cooling. See our guide.
Hyperscaler: a very large cloud provider operating massive data centers.
PUE: a measure of data center energy efficiency. See cooling and power.
Vector database: stores data as numbers so AI can find related information. See data storage.
Qubit: the basic unit of quantum computing. See quantum AI.
Data
Data science: extracting knowledge and predictions from data. See our guide.
Data analysis: examining data to answer questions. See our guide.
Dataset: a collection of data used for training or analysis.
Knowledge cutoff: the date after which a model has no trained knowledge.
How PositoAI fits in
PositoAI is a free, AI-powered stock information, education, research and screening assistant platform. Learn more on our FAQ and explore the full AI series on our blog.
Summary
This glossary explains key AI terms, from basics like machine learning and LLMs, to usage terms like prompts, agents and RAG, risk terms like bias, deepfakes and alignment, and infrastructure terms like GPUs, HBM, data centers and PUE. AI vocabulary changes quickly, so this page will be updated over time.