Artificial intelligence (AI) is technology that lets computers perform tasks that normally require human thinking, such as understanding language, recognizing images, spotting patterns, and making predictions. Instead of following only fixed, hand-written rules, most modern AI systems learn from large amounts of data and improve at a task over time.
AI is now part of everyday computing. It powers the chatbot that drafts your email, the camera that keeps you in frame on a video call, and the laptop feature that blurs your background or reduces noise from a barking dog. Newer AI PCs even include dedicated hardware built to run these features directly on the device.
This AI glossary explains the terms you are most likely to see when you shop for a computer, use AI tools at work, or read about AI in the news. Each topic below links to a detailed guide, and a quick-reference list of key terms follows for fast answers.
Explore the AI glossary by topic
The guides below are grouped by what you want to understand: the core technology, the hardware that runs it, how you use it, and how to use it responsibly.
Core AI concepts
Machine learning Machine learning is the branch of AI that teaches computers to learn patterns from data rather than following only programmed rules. This guide covers how models are trained, the main types of machine learning, and where you encounter them every day.
Generative AI Generative AI creates new content, including text, images, audio, video, and code, based on patterns learned from existing examples. Learn how it works, what it is good at, and where its limits are.
Large language models (LLMs) A large language model is a type of generative AI trained on vast amounts of text to understand and produce human language. This guide explains tokens, prompts, context windows, and why LLMs sometimes get facts wrong.
AI hardware and computing
AI PCs An AI PC is a computer designed to run AI workloads locally, usually with a CPU, GPU, and neural processing unit working together. Find out what separates an AI PC from a standard laptop and what to look for when you buy one.
Neural processing units (NPUs) An NPU is a processor built specifically to run AI tasks efficiently while using less power than a CPU or GPU. This guide covers how NPUs work, what TOPS ratings mean, and which features rely on them.
On-device AI vs. cloud AI On-device AI runs directly on your computer, while cloud AI sends your request to remote servers for processing. Compare the trade-offs in speed, privacy, cost, and capability, and learn when each approach makes sense.
Using AI
AI assistants AI assistants are tools that understand natural-language requests and help you complete tasks, from answering questions to summarizing documents. Learn how they work and how to get more useful results from them.
AI features AI features are the specific capabilities built into devices and software, such as background blur, noise cancellation, auto-framing, and smart battery management. This guide explains common AI features and what they do for you.
AI for business Businesses use AI to automate routine work, analyze data faster, and support employees and customers. Explore common business uses, practical benefits, and what teams should consider before adopting AI tools.
Responsible AI
AI privacy and security AI tools often handle personal and business data, which raises questions about how that data is stored, used, and protected. This guide covers key privacy and security risks and the steps you can take to reduce them.
Key AI terms at a glance
Use this A–Z list for a quick definition, then follow the link for the full explanation.
Term | Definition | Learn more |
|---|---|---|
AI agent | An AI system that can plan and carry out multi-step tasks, such as researching a topic or filling out a form, with limited human input. | |
AI assistant | Software that understands natural-language requests and helps you complete tasks like answering questions, drafting text, or summarizing files. | |
AI PC | A computer with hardware, typically including an NPU, designed to run AI tasks locally and efficiently. | |
Algorithm | A set of step-by-step instructions a computer follows to solve a problem or complete a task. | |
Artificial intelligence (AI) | Technology that enables computers to perform tasks that normally require human intelligence, such as understanding language or recognizing images. | This page |
Chatbot | A program that holds a conversation with users through text or voice, often powered by a large language model. | |
Cloud AI | AI that runs on remote servers, so your request is sent over the internet and the result is returned to your device. | |
Computer vision | A field of AI that lets computers interpret images and video, used in features like auto-framing and face detection. | |
Context window | The amount of text, measured in tokens, that a language model can consider at one time. | |
Deep learning | A type of machine learning that uses neural networks with many layers to learn complex patterns. | |
Generative AI | AI that creates new content, such as text, images, audio, or code, based on patterns in its training data. | |
GPU | A graphics processing unit, a processor that handles many calculations at once and is widely used to train and run AI models. | |
Hallucination | When an AI model produces information that sounds confident but is false or made up. | |
Inference | The process of a trained AI model using what it has learned to produce an answer or prediction. | |
Large language model (LLM) | A generative AI model trained on vast amounts of text to understand and generate human language. | |
Machine learning (ML) | A branch of AI in which computers learn patterns from data instead of following only explicit rules. | |
Model | The trained system that takes an input, such as a question or an image, and produces an output. | |
Natural language processing (NLP) | The field of AI focused on helping computers understand, interpret, and generate human language. | |
Neural network | A computing system loosely modeled on the brain, made of connected layers that process data and learn from it. | |
NPU | A neural processing unit, a processor designed to run AI workloads efficiently with low power use. | |
On-device AI | AI that runs locally on your computer or phone rather than on remote servers. | |
Prompt | The instruction or question you give an AI tool to tell it what you want. | |
TOPS | Trillions of operations per second, a measure of how many AI calculations a processor such as an NPU can perform. | |
Token | A small unit of text, such as a word or part of a word, that a language model reads and generates. | |
Training data | The information an AI model learns from, such as text, images, or numbers. |
How these AI terms fit together
Many AI terms describe layers of the same idea. Large language models are a type of generative AI, generative AI is built on deep learning, deep learning is a form of machine learning, and machine learning is one approach to AI. Separately, hardware terms like NPU and AI PC describe where that AI actually runs.
A single feature can draw on several layers at once. For example, an AI assistant on an AI PC might use a language model running on the NPU for quick tasks and send larger requests to the cloud.
Frequently asked questions about AI
What is the difference between AI, machine learning, and deep learning?
AI is the broad field of making computers perform tasks that require human-like intelligence. Machine learning is one way to achieve AI, where systems learn from data. Deep learning is a type of machine learning that uses many-layered neural networks, and it powers most of today's advanced AI, including image recognition and language models.
What is the difference between AI and generative AI?
Generative AI is a subset of AI that creates new content. Traditional AI typically analyzes or classifies existing information, such as flagging spam or recommending a movie. Generative AI produces something new, such as a written summary, an image, or a block of code. Learn more in our generative AI guide.
How does AI learn?
Most AI learns through a process called training, where a model studies large amounts of example data and adjusts itself to find patterns. Once trained, the model uses those patterns to respond to new inputs, a step called inference. Our machine learning guide explains the process in more detail.
What is an AI PC, and do I need one?
An AI PC is a computer with dedicated hardware, usually a neural processing unit, that runs AI features locally and efficiently. You do not need one to use web-based AI tools like chatbots. An AI PC is worth considering if you want AI features that work offline, respond faster, keep more data on your device, and use less battery. See our AI PC guide for what to look for.
What does an NPU do in a laptop?
An NPU handles AI tasks, such as background blur, voice isolation, and live captions, so the CPU and GPU are free for other work. Because NPUs are built for AI math, they can run these tasks using much less power, which helps preserve battery life. NPU performance is measured in TOPS. Microsoft's Copilot+ PC standard, for example, requires an NPU rated at 40 or more TOPS. Read more in our NPU guide.
Is on-device AI more private than cloud AI?
On-device AI can be more private because your data is processed on your computer instead of being sent to remote servers. Cloud AI can access larger, more powerful models but requires sending your request over the internet. Many tools use a mix of both. Compare them in our on-device vs. cloud AI guide.
Can AI make mistakes?
Yes. AI models can produce wrong answers, misread context, or reflect biases in their training data. Language models can also "hallucinate," stating made-up facts with confidence. Always review AI-generated content before you rely on it, especially for important decisions.
Is it safe to use AI tools?
AI tools are generally safe when you use reputable providers and follow good data habits. Avoid entering passwords, financial details, or confidential business information unless you know how the tool stores and uses your data. Our AI privacy and security guide covers the main risks and how to reduce them.
How are businesses using AI?
Businesses use AI to automate repetitive tasks, summarize meetings and documents, analyze data, support customer service, and help employees write and research faster. The most successful teams start with a specific problem and set clear rules for data use. Explore examples in our AI for business guide.
How can I start using AI?
The easiest way to start is with an AI assistant built into tools you already use, such as your operating system, browser, or office apps. Try asking it to summarize a long document, draft an email, or brainstorm ideas. Our AI assistants guide shares tips for writing prompts that get better results.
