Everyday AI vocabulary
Start with AI basics. 28 hand-written terms with real workplace examples.
Learn everyday ai terms every day.
One hand-written term. A real workplace example. Thirty seconds before your day gets loud.
AI Model
The engine inside an AI tool. It learned patterns from a huge number of examples, like books, websites and photos, and uses them to answer what you ask.
Large Language Model
The kind of AI model behind most AI chat tools, often called an LLM. It learned from huge amounts of written text, and it writes by predicting likely next words, a piece at a time.
Generative AI
AI that makes something new when you ask, like a letter, a picture, a song or a voice. Earlier everyday AI mostly sorted or flagged things, like a spam filter; this kind creates.
Training Data
The huge collection of examples, like books, websites, photos and recordings, that an AI model learned its patterns from. What it includes, and what it leaves out, shapes every answer.
Chatbot
A program you talk with by typing or speaking, and it answers back in a conversation. Today's AI versions run on a large language model, so they write new replies instead of reading from a script.
Prompt
The message you type or say to an AI tool to ask for something. Clear ones say what you want, who it is for and any details that matter, like length or reading level.
Token
A small chunk of text, often a short word or a piece of a longer one, that a large language model reads and writes. AI tools count length and limits in tokens; these are not login or access tokens.
Follow-Up
A next message that builds on the AI's last answer instead of starting over, like 'shorter,' 'simpler words' or 'add the date.' A few quick ones usually get you further than one perfect first try.
Fact-Check
Confirming names, dates, numbers and claims against a trusted source before you use them. AI can sound sure and still be wrong, so check anything that matters.
Source Citation
A note that shows where a claim came from, like a link, a book page or a report name, so you can open it and check. A link alone is not proof; make sure the page really says that.
Knowledge Cutoff
The date after which an AI model's training data stops. It may not know about anything newer, like a new law or this week's news, unless the tool can search the web.
Human Review
A person reading and checking AI work before it goes out or anyone relies on it. It catches wrong facts, an awkward tone and missing details that slip past the AI.
AI Slop
Low-quality digital content made in quantity with AI, like generic posts, odd-looking images and filler videos. The problem is careless mass production, not every use of AI.
Workslop
AI-made work, like a memo, report or slide deck, that looks polished but lacks the substance to move the task forward, so a coworker has to fix or redo it. Office-sized slop.
Deepfake
A fake video, photo or voice recording made with AI to look or sound like a real person. Scammers can clone a voice from a short clip and call family members asking for money.
AI Disclosure
A short, honest note that AI helped make something, like 'Drafted with AI, checked by our staff.' Many schools and some platforms ask for one, and it builds trust either way.
Sensitive Data
Information that could hurt someone if it leaked, like health details, Social Security numbers, passwords, student records or donors' bank details. Keep it out of AI tools your organization hasn't approved for it.
Training Opt-Out
A setting, in tools that offer one, that stops your chats from being used to train future AI models. Look under privacy or data settings; it does not delete chats you already had.
AI Memory
A feature that lets an AI assistant remember details about you across separate chats, like your kids' ages or your job. Where it's offered, you can usually view, turn off or delete what it saved.
Shadow AI
Staff using AI tools their organization hasn't approved, often with work data. It can put sensitive data at risk, so ask which tools are allowed or help get a good one approved.
Internal Workflow
The behind-the-scenes steps your team repeats to get regular work done, like logging donations or preparing the monthly report. A good first place to try AI is one step, with a person checking it.
AI Summary
A short version of a longer document, recording or email thread, written by AI so you get the main points fast. It can drop or bend details, so check the original before acting on anything important.
Transcription
Turning spoken words into written text, like notes from a meeting recording or a voicemail. AI does it fast but can mishear names and numbers, so ask before recording people and check the text before sharing.
AI Search
Asking a question in everyday words and getting an answer drawn from a set of documents, like your team's shared files or the web, often with links to where it found each part.
Custom Assistant
An AI assistant you set up once with standing instructions and your own reference files, so it does one repeated job the same way every time, like drafting the weekly bulletin.
Template
A reusable pattern with blanks to fill in, like a saved prompt for weekly reminders. You keep what stays the same and change only the details, so you never start from scratch.
Share Link
A web link that lets other people open a chat or something you made, without sending a file. Anyone with the link can usually see it, so check what's inside first and delete the link when you're done.
No-Code
Building a small tool, like a sign-up form, a quiz or a simple app, by describing it, clicking and dragging instead of writing computer code. AI now lets you build some of these just by asking.
Building toward a role? Browse career pathways that map terms to jobs.