The Plain-English AI Glossary

AI (Artificial Intelligence)

Computer systems that do things we used to think needed a human brain — understanding language, answering questions, spotting patterns, making suggestions.

Think of it asA very well-read assistant who has read almost everything but has never actually lived in the world.

Why it matters to youIt's the umbrella term for all the tools you're learning. When someone says “use AI,” they usually mean a tool like Claude or ChatGPT.

You'll see this whenEverywhere — “AI-powered,” “add AI to this,” “ask the AI.”

Generative AI

A type of AI that creates new things — text, images, summaries — rather than just sorting or searching what already exists.

Think of it asThe difference between a librarian who points you to a book and a writer who drafts a new one on the spot.

Why it matters to youThis is the kind you'll use daily to draft emails, summarize calls, and brainstorm. “Generative” is why it can produce a first draft from a blank page.

You'll see this when“GenAI,” “generative tools,” anytime AI writes or makes something for you.

Large Language Model (LLM)

The engine behind text-based AI — a program trained on enormous amounts of writing so it can predict and produce natural-sounding language.

Think of it asAutocomplete on your phone, but vastly more powerful — it can finish not just your word, but your whole document.

Why it matters to youClaude, ChatGPT, and Gemini are all LLMs. Knowing this explains both their power and their limits: they predict likely words, they don't truly “know” facts.

You'll see this when“The model,” “LLM,” “language model.”

Model

A specific AI “brain” you're talking to, each with a name and version (Claude Opus, GPT-4, Gemini). Newer ones are generally smarter.

Think of it asDifferent editions of a car — same idea, but a newer model handles better and has more features.

Why it matters to youYou sometimes get to choose which model to use; a stronger one gives better results on hard tasks. It also explains why answers differ between tools.

You'll see this whenA dropdown that says “Model: …,” or “which model are you using?”

Training data

All the text and information an AI learned from before you ever talked to it. It's how the model “knows” things.

Think of it asEverything the assistant read during school, before starting the job with you.

Why it matters to youThe AI's general knowledge comes from here — which is also why it can be out of date or occasionally wrong.

You'll see this when“It was trained on…,” “that's from its training.”

Knowledge cutoff

The date after which the AI has no built-in knowledge. It doesn't automatically know things that happened later.

Think of it asAn assistant who's been on a desert island since a certain date — sharp as ever, but hasn't seen today's news.

Why it matters to youFor anything recent — prices, news, who holds a role — give the AI current info or have it search the web. Don't trust its memory for fresh facts.

You'll see this when“As of my knowledge cutoff,” or when it gets a recent fact wrong.

Hallucination

When AI states something false as if it were true — a made-up fact, fake quote, or invented source — usually said with total confidence.

Think of it asA confident intern who'd rather give you an answer than admit “I don't know.”

Why it matters to youThis is the #1 reason to verify anything important. The AI isn't lying on purpose — it fills gaps with plausible-sounding guesses.

You'll see this whenA citation that doesn't exist, a stat that's a little too perfect, a “fact” you can't confirm.

Prompt

The instruction or question you give the AI — anything you type or say to it.

Think of it asThe brief you hand a contractor. A vague brief gets vague work; a clear one gets what you wanted.

Why it matters to youThe quality of your prompt is the single biggest lever on the quality of your result. Better prompts = better output, every time.

You'll see this whenThe box you type into; “write a better prompt,” “prompt it to…”

Prompt engineering

The skill of writing prompts that get great results — being clear, giving context, showing examples, and asking for the format you want.

Think of it asLearning to give directions well, so people actually end up where you meant.

Why it matters to youIt's a learnable skill, not a talent. Getting good at it is most of what separates “vending machine” from “thinking partner.”

You'll see this when“Prompt engineering,” “prompting techniques,” “prompt tips.”

Experience prompting

(Our term.) Prompting by describing the experience and outcome you want — the situation, who it's for, what “good” looks like — instead of barking a one-line command.

Think of it as“Make coffee” vs. “I've got a stressed client arriving in 5 minutes who loves a strong oat-milk latte.” The second gets you something genuinely useful.

Why it matters to youThis is the core habit we teach — how you turn the AI into a partner that understands your world, not a machine that guesses.

You'll see this whenThroughout this class — it's the “how.”

Context

The background you give the AI so it understands your specific situation — who you are, the client, the goal, the constraints.

Think of it asThe difference between a stranger and a coworker who already knows your business. Context makes the AI a coworker.

Why it matters to youMore relevant context = dramatically better answers. Most weak results come from too little context, not a weak AI.

You'll see this when“Give it more context,” “for context, …”

Context window

How much the AI can “hold in mind” at once — your messages + its replies + any files, all together in one conversation.

Think of it asThe size of the desk it's working on. A bigger desk holds more papers; when it's full, older papers fall off the edge.

Why it matters to youIn very long chats it may “forget” things from far earlier. If that happens, remind it or start fresh with the key info.

You'll see this when“It lost track,” “running out of context,” long chats getting fuzzy.

Token

The small chunks of text AI reads and writes in — roughly a word or piece of a word. It's how AI “counts” length.

Think of it asSyllables. The AI reads in syllable-sized pieces rather than whole words or letters.

Why it matters to youYou mostly don't need to worry about it, but it explains limits (“too long”) and, on paid plans, cost — usage is measured in tokens.

You'll see this when“Token limit,” “you've used X tokens.”

System prompt / instructions

Behind-the-scenes instructions that set how the AI behaves for a whole project — its standing rules, before you type anything.

Think of it asThe job description and house rules you give a new hire on day one, so you don't repeat yourself every task.

Why it matters to youGood instructions (like “always read theBrain first”) make every future answer better automatically. Set once, benefit forever.

You'll see this when“Project instructions,” “custom instructions,” “system prompt.”

Thread / conversation

A single ongoing chat with the AI. Everything in one thread shares context; a new thread starts fresh.

Think of it asOne email chain vs. a brand-new email. Reply keeps the history; a new message starts blank.

Why it matters to youKeep related work in one thread so the AI remembers it; start a new one when you switch topics to avoid confusion.

You'll see this whenYour list of past chats; “start a new chat,” “in this thread.”

Iteration (following up)

Improving a result by responding to it — “make it shorter,” “warmer tone,” “add a line on pricing” — instead of accepting the first draft.

Think of it asWorking with a designer through rounds of feedback, not expecting perfection on version one.

Why it matters to youThe magic is usually in round two and three. Treating the first answer as a draft, not a verdict, is a thinking-partner habit.

You'll see this whenAny time you reply to refine instead of retyping the whole request.

Examples (few-shot)

Showing the AI one or two samples of what you want, so it matches the pattern.

Think of it asHanding someone a sample before saying “make three more like this.”

Why it matters to youExamples are one of the fastest ways to lift quality — especially for formatting, tone, or a repeatable task.

You'll see this when“Here's an example of a good one…,” “match this style.”

AI as a vending machine

Using AI for quick, one-shot requests — put in a question, get an answer, walk away. Transactional and shallow.

Think of it asA vending machine: push a button, get a snack. Fine for a snack, useless for a meal.

Why it matters to youIt's where most people start — nothing wrong with it — but it leaves most of the value on the table. Recognizing this is step one.

You'll see this whenOne-line questions, copy-paste, never following up.

AI as a thinking partner

Working with AI — giving it context, going back and forth, letting it help you think, plan, and create, not just fetch.

Think of it asA sharp colleague you brainstorm with over coffee, not a machine you bark orders at.

Why it matters to youThis is the whole point of the class. It's where AI starts making you genuinely more valuable at your job.

You'll see this whenLonger conversations, sharing context, asking “what am I missing?”

Zone of Genius

(From Gay Hendricks.) The work you're uniquely great at and that energizes you — as opposed to what you're merely good at, or just competent at.

Think of it asThe part of your job that feels like play but looks like magic to everyone else.

Why it matters to youUsing AI well means offloading the draining tasks so you spend more time here — that's how you become more valuable, not replaceable.

You'll see this whenYour Day-1 AI Mastery Profile; “what only you can do.”

Delegating vs. collaborating

Delegating = handing off a whole task and hoping. Collaborating = working alongside the AI, steering as you go.

Think of it asMailing a letter and waiting vs. having a live conversation.

Why it matters to youFor anything that matters, collaborating beats delegating — you catch issues early and shape the result.

You'll see this whenThe difference between “do this for me” and “let's work through this.”

Claude

The AI assistant made by Anthropic — the one powering this class. Known for thoughtful writing, long documents, and working with your files.

Think of it asOne brand of AI assistant, like one brand of phone. Same category as ChatGPT and Gemini, different maker.

Why it matters to youIt's the tool you'll build your brain and profile in. Everything in this plugin runs on Claude.

You'll see this whenThe app you're in; claude.ai.

Cowork

A mode inside the Claude app where the AI does multi-step work for you — uses your files, runs tasks, and produces finished documents — not just chat.

Think of it asThe difference between asking a coworker a question and handing them a whole project to complete.

Why it matters to youIt's where the real work happens — building deliverables, running skills, saving files — while you stay in the loop.

You'll see this when“Cowork mode,” tasks that produce actual files.

Artifact

A finished piece of work the AI creates that appears in its own panel — a document, web page, chart, or checklist you can view, keep, and reuse.

Think of it asThe deliverable on the table, not the conversation about it.

Why it matters to youYour polished outputs (like this glossary) become artifacts you can revisit and share, instead of getting lost in chat history.

You'll see this whenA side panel opening with a document or page; “saved as an artifact.”

Project

A dedicated workspace in Claude for one area of work, with its own files and standing instructions, so the AI has consistent context every time.

Think of it asA labeled folder-plus-desk for one client or job, set up exactly how you like it.

Why it matters to youProjects are where your brain lives and gets read automatically — set it up once, and every chat inside it is already up to speed.

You'll see this when“Create a project,” “project instructions.”

Plugin

An installable bundle that adds ready-made abilities (skills and commands) to Claude in one click.

Think of it asAn app you install on your phone to add a new capability.

Why it matters to you“IMG · Become AI Native” is a plugin — installing it gives your whole team the same tools at once.

You'll see this when“Install the plugin,” the plugin bundle for this class.

Skill

A packaged set of expert instructions that teaches the AI how to do a specific task really well, every time.

Think of it asA trained specialist you can call in — the AI follows a proven recipe instead of winging it.

Why it matters to youSkills make results consistent and high-quality. You often don't call them directly; a command runs them for you.

You'll see this when“This skill does…,” the skills list in your plugin.

Command (slash command)

A shortcut you type starting with “/” to trigger a specific action — like /setup or /draft.

Think of it asA speed-dial button for a whole task.

Why it matters to youCommands are the easy front door to the plugin's power — no need to remember long prompts.

You'll see this whenTyping “/” and seeing a menu; /ai-profile, /meeting-prep.

Connector / MCP

A secure link between the AI and another tool you use (Gmail, Slack, Drive, your CRM) so it can read or act there with your permission. MCP is the technology behind those links.

Think of it asA power adapter that lets your AI plug into the apps you already use.

Why it matters to youConnectors are what let the AI prep a meeting from your calendar or draft from your email — real work, in your real tools.

You'll see this when“Connect your tools,” “connector,” a permission pop-up.

Wispr Flow

A voice-to-text tool that types what you say — fast and accurately — anywhere on your computer.

Think of it asA tireless transcriptionist sitting at your keyboard.

Why it matters to youSpeaking is faster than typing and gives the AI richer context. We use it so you can talk to your AI, especially for your AI Mastery Profile.

You'll see this whenThe class tools setup; dictating instead of typing.

Chatbot vs. Agent (agentic AI)

A chatbot answers what you ask. An agent takes actions and completes multi-step tasks for you — deciding steps, using tools, and reporting back.

Think of it asA receptionist who answers questions vs. an assistant who actually runs the errand.

Why it matters to youThe tools are shifting from answering to doing. Understanding “agentic” helps you hand off bigger jobs safely.

You'll see this when“Agent,” “agentic,” AI that does things, not just talks.

Markdown / Markdown file (.md)

A simple way to format plain text — headings, bold, bullets, links — using a few keyboard symbols instead of buttons. A file saved this way ends in .md.

Think of it asTexting shorthand for documents — you type **bold** and it becomes bold.

Why it matters to youAI often replies in Markdown, and “brain” files are Markdown files. Recognizing it stops it from looking like scary code.

You'll see this when# symbols, **asterisks**, or a file ending in .md.

Plain text vs. rich text

Plain text is words with no formatting (like Notepad). Rich text has fonts, colors, sizes, and styling (like Word).

Think of it asA handwritten note vs. a designed flyer.

Why it matters to youAI and brain files love plain text/Markdown because it's clean and portable; you convert to rich formats (Word, PDF) for the final, pretty version.

You'll see this when.txt vs .docx; “paste as plain text.”

File formats (.docx, .pdf, .csv, .xlsx, .html)

The “type” of a file, shown by the letters after the dot. Each opens in different apps and is good for different jobs.

Think of it asContainers — a mug, a bottle, a box — each shaped for what goes in it.

Why it matters to youKnowing them tells you what a file is: .docx = Word doc, .pdf = fixed printable, .csv/.xlsx = spreadsheets, .html = web page.

You'll see this whenThe end of every filename.

CSV

A plain-text spreadsheet — “Comma-Separated Values.” Rows of data with commas between the columns. Opens in Excel or Google Sheets.

Think of it asA spreadsheet stripped to its bare bones — just the data, no colors or formulas.

Why it matters to youIt's the universal format for moving lists and data between tools (contacts, exports, imports). AI often hands data back as CSV.

You'll see this whenA file ending in .csv; “export as CSV.”

Metadata

Data about data — the hidden details attached to a file, like who created it, when, and its title.

Think of it asThe label on a soup can — not the soup, but everything about it.

Why it matters to youGood to know files carry hidden info (dates, authors), especially when sharing documents with clients.

You'll see this whenFile “properties” or “Get Info”; document details.

theBrain

Your company's single source of truth, written for the AI — your voice, services, ideal client, story, rules, and goals — so the AI works from your reality.

Think of it asThe company handbook and brand guide, but written so the AI reads it before every task.

Why it matters to youIt's the foundation of everything here. With theBrain, the AI stops guessing and starts sounding like your business.

You'll see this when/setup; the brain.md file and its sections.

roleBrain

The same idea as theBrain, but for your specific job — your SOPs, tools, tasks, hand-offs, and boundaries.

Think of it asYour personal job binder, readable by the AI.

Why it matters to youIt makes the AI genuinely useful for your seat, and it plugs into the company theBrain.

You'll see this when/build-role-brain; your roleBrain folder.

SOP (Standard Operating Procedure)

A written, step-by-step way of doing a recurring task, so it's done the same way every time.

Think of it asA recipe — anyone (or any AI) can follow it and get the same dish.

Why it matters to youYour SOPs go into your roleBrain so the AI can help run your processes exactly how you do them.

You'll see this when“Document your SOP,” “the process for X.”

Knowledge base

An organized, central place where information is stored so it can be found and reused — a step above scattered notes.

Think of it asA well-organized filing cabinet everyone (including the AI) can pull from.

Why it matters to youtheBrain and roleBrain together are your AI-ready knowledge base — that's what makes the AI consistent.

You'll see this when“KB,” “knowledge base,” a help center.

Source of truth

The one place you trust as correct when information conflicts — the master copy.

Think of it asThe original signed contract vs. five slightly different photocopies floating around.

Why it matters to youtheBrain is the AI's source of truth. When something's unclear, that's where the right answer lives.

You'll see this when“Single source of truth,” “which version is correct?”

Claim-rules / guardrails

Rules that limit what the AI is allowed to say — especially for client-facing or regulated content (e.g., no medical claims you can't back up).

Think of it asGuardrails on a mountain road — they keep you from going off a cliff.

Why it matters to youThey protect you and your clients from saying something inaccurate or non-compliant. A flagged rule is a hard stop.

You'll see this when“Claim-rules,” “guardrails,” ⚠️ flags in output.

Verify (“trust but verify”)

Checking the AI's important claims against a reliable source before you use them.

Think of it asProofreading an intern's work before it goes to the client.

Why it matters to youBecause AI can hallucinate, verifying facts, figures, names, and links is non-negotiable for anything client-facing.

You'll see this whenDouble-checking a stat, a name, a date, a link.

Data privacy / confidentiality

Being careful about what sensitive information you put into AI tools — client data, passwords, private details.

Think of it asYou wouldn't leave client files open on a café table; treat AI inputs the same way.

Why it matters to youNever paste passwords or one-time codes; be thoughtful with client data and follow your company's rules. Brain files record where things live, never the passwords themselves.

You'll see this when“Is this safe to share?”, your company data policy.

Bias

When AI output leans a certain way or reflects skewed patterns from its training data — sometimes subtly.

Think of it asA newspaper with a slant — useful, but you read it knowing the angle.

Why it matters to youStay aware that AI can reflect stereotypes or one-sided views; your human judgment is the check.

You'll see this whenOutput that assumes or generalizes; “that seems biased.”

Workflow / automation

A workflow is the series of steps to get a job done; automation is having the AI or software do some of those steps for you, repeatedly.

Think of it asAn assembly line for a routine task — set it up once, it runs each time.

Why it matters to youThe big wins come from automating your recurring work, not one-offs. That's the Day-2 build.

You'll see this when“Automate this,” “set up a workflow.”

Template / prompt library

Saved, reusable prompts or document formats for tasks you do often, so you don't start from scratch.

Think of it asCanned email responses — write it once, reuse it forever.

Why it matters to youBuilding a small library of your best prompts is one of the highest-return habits you can build.

You'll see this when“Save this as a template,” “my prompt library.”