The AIEB | AI Glossary
Playbook
October 1, 2026
6 min read
On this page
Prefer to print it? Download this glossary as a PDF. It's the same content, laid out for printing.
New to AI? You're going to hear a lot of new words in this bootcamp. This page explains every one of them in plain English.
You don't need to memorize anything. Come back here any time a word trips you up.
💡 The basics
- AI: software that can understand plain language and write, analyze or create things the way a person would.
- Chat: a conversation with an AI. You ask, it answers. Each new chat starts from zero.
- Prompt: what you type or say to the AI to tell it what you want.
- Model: the "brain" inside an AI tool. Companies release newer, smarter models every few months.
- Context: everything the AI knows about your situation when it answers. More good context means better answers.
- Context window: how much the AI can keep in mind at once. In a very long chat, older details can get pushed out, so start a fresh chat when things get messy.
- Hallucination: when an AI states something false with confidence. That's why you always check its work.
- Transcript: a written, word-for-word record of a call, meeting or voice memo. It's one of the most useful things you can give an AI.
- Token: a small chunk of text, about three-quarters of a word. AI tools measure what you send and get back in tokens.
- Usage limits: how much you can use an AI tool before it asks you to wait or upgrade. Higher plans give you more.
- Web search: when the AI looks things up online, so its answer includes current information.
- Voice mode: talking to the AI out loud and hearing it answer back, like a phone call.
- Dictation: speaking instead of typing, so your words turn into text. Wispr Flow is one popular app for this.
- Extended thinking and deep research: settings that let the AI take more time to think or research before it answers, for harder questions.
🧩 The parts of an AI employee
- AI employee: an AI you've trained to do specific jobs in your business, consistently, without being asked twice.
- Skill: a saved set of instructions for one job. The AI uses it the same way every time.
- Skill Creator: a built-in Claude skill that helps you build new skills.
- Project (or workspace): where an AI employee lives. It holds your business files, instructions and memory for one part of your business.
- Project knowledge: the files you upload to a project, like your offers, your ideal client and your best past work.
- Knowledge base: a collection of your business documents the AI can learn from. In a Claude Project, that's your project knowledge.
- SOP (standard operating procedure): a written, step-by-step description of how a task gets done in your business. SOPs are the raw material for great skills.
- ICP (ideal client profile): a clear description of the exact customer you serve best. It's one of the most important things to give your AI.
- Instructions: the standing rules for a project. They tell the AI who it is, who it's helping and how it should sound.
- Memory: what the AI remembers from past conversations, so you don't have to repeat yourself.
- Connector: a link that lets the AI read and act inside your other apps, like Gmail, Slack or Notion. ChatGPT calls these apps.
- MCP: the technical standard that connectors are built on. If an app "has an MCP server," it can connect to AI.
- Scheduled task: a job the AI runs on its own at a set time, like a report every Monday at 7am.
- Business brain: one organized home for everything your business knows, which your AI employees can learn from.
- Second brain: your personal version of a business brain, with your ideas, notes and decisions in one place.
Want to see how these parts fit together? See The AIEB | What Is an AI Employee? for the four levels, from AI Contractor to Autonomous AI Employee.
🎨 What the AI makes
- Artifact: a finished thing the AI makes for you, like a document, a dashboard or a small app.
- Image generation: asking the AI to create a picture from a description. This is what we mainly use ChatGPT for.
- Live artifact: an artifact that pulls in fresh information from your apps, so it stays up to date.
- Claude Docs and Claude Design: Claude's built-in tools for making polished documents and visual designs.
- .md file: a simple text file format (short for "markdown") that AI reads easily. Many skills and instructions are saved this way.
🛠️ The tools
- Claude: the AI assistant made by Anthropic, and our main tool.
- ChatGPT: the AI assistant made by OpenAI, and our go-to for images.
- ChatGPT Work: a mode in ChatGPT for handing off bigger jobs and getting finished files back.
- Custom GPT: an older way to save a customized assistant in ChatGPT. It's being replaced by skills and projects.
- Cowork: Claude's former name for handing off bigger jobs. It's now built right into Claude.
- Claude Code and Codex: tools built for large, complex jobs, made with more technical users in mind. Claude Code is from Anthropic, and Codex is from OpenAI.
- Claude in Chrome: a Claude add-on for your web browser that can read pages, click and fill in forms for you.
- Plugin: in ChatGPT, a package that adds skills or app connections.
- NotebookLM: Google's AI tool for learning. Upload your sources and it teaches you the topic.
- Perplexity: an AI tool built for research that shows you where its answers come from.
- Granola: an app that records and transcribes your meetings. It's what we use.
- Replit: a tool for building apps and websites by describing them to AI.
- Muse and GrokBot: newer AI assistants for small, personal tasks. Muse is Meta's, and GrokBot is built on Grok, from Elon Musk's company xAI.
- Apify: a tool that collects public information from websites and social media for your AI to use.
🧭 Ways of working
- AI User: someone who pastes questions into a chat and starts over every time.
- AI Architect: someone who builds and trains AI employees that keep doing the work.
- Agent: AI that can take steps on its own instead of only answering questions.
- Human in the loop: keeping a person involved to check or approve the AI's work before it goes out.
- Computer use: when the AI controls a computer or browser to click, type and move through apps for you.
- SEO: search engine optimization, which means getting found on Google.
- GEO: generative engine optimization, which means getting found and recommended inside AI answers, like ChatGPT and Claude.
- Vibe coding: building an app by describing what you want in plain English and letting AI write the code.
Not in A2A? Head back to the Bootcamp Hub → and enter the email you purchased with, then click "Send My Sign-In Link." It has to be the same email you bought with.
In A2A? Log in to TUB Connect with your A2A email.
