AAI Word
Bank

Explore

AI Word Bank

AI vocabulary, explained through a familiar kitchen.

A beginner’s tasting menu

AI vocabulary.
One idea at a time.

You don’t need to memorize every technical word to start using AI. Learn these essentials in order, then open any term for its definition and illustrated kitchen example.

Looking for a specific word? Search all 128 terms →

01

Meet the kitchen

Start with the difference between the field, the trained system, the app you open, and the material it learns from.

Artificial intelligence (AI)

When computers are designed to learn, decide, create, or solve problems that normally require human intelligence.

AI model

A mathematical system trained on data to perform tasks such as writing, translating, or classifying.

AI tool

An application or service that lets people use an AI model to complete a task.

Dataset

An organized collection of examples—such as text, images, or numbers—used to train or test AI.

02

Place a clear order

These words explain how you ask for help and what the model can work with while answering.

Prompt

The message, question, or instructions you give an AI model to explain what you want.

Output

The answer or content an AI returns after receiving a prompt.

Token

A small piece of text—a word or part of a word—that a language model reads and generates.

Context window

The maximum amount of information a model can consider at one time.

03

Understand how it learns

Learning from examples and answering a new request happen at different stages. These terms help you tell them apart.

Machine learning (ML)

A way for computers to learn patterns from data instead of following only hand-written rules.

Training

The phase when an AI model studies examples and adjusts itself to make better predictions.

Large language model (LLM)

A model trained on huge amounts of text so it can predict language and generate fluent responses.

Inference

The moment a trained model uses what it learned to answer a new question or make a prediction.

04

Check before serving

A confident response still needs judgment. Learn how sources, clear context, and human review can improve the result.

Hallucination

An AI response that sounds confident but contains incorrect, misleading, or invented information.

AI Grounding

Giving an AI trusted, relevant information such as documents, database records, or search results so its answer stays connected to real sources.

AI slop

Low-quality text, images, or video made with AI and published mainly for volume or attention.

Human-in-the-loop

Keeping a person involved to guide, approve, or correct an AI system’s work.

05

Give the kitchen tools

Move from asking for an answer to planning work, connecting software, and setting limits on what an AI can do.

AI agent

An AI system that can plan steps, use tools, and take actions toward a goal.

API (Application Programming Interface)

A defined way for software systems to send requests and exchange information.

Model Context Protocol (MCP)

An open standard that lets an AI app connect to outside tools, data, and services through one consistent setup.

Agent authorization

Checking whether a particular AI agent may take a particular action on a particular resource under the current rules.

Three useful distinctions

Similar words, different jobs.

A model and a tool

The model is the trained system that produces a response. The tool is the application around it, with an interface and features you use. Think of a chef and the restaurant that serves the meal.

Training and inference

Training adjusts a model using examples. Inference uses the trained model to handle a new request. A chef practicing recipes and a chef preparing your order are doing different jobs.

A chatbot and an agent

A chatbot focuses on conversation. An agent can also plan steps and use tools to act toward a goal. Capabilities can overlap, so check what a particular system can actually do and what it is allowed to change.