Glossary of AI terms

Because “natural language processing” shouldn’t make you want to nap.

 
 

Welcome to the least boring AI glossary you’ll ever read.

If you’ve ever nodded along to tech talk while secretly thinking, “What even is a prompt?”—this is for you.

We’ve broken down the buzzwords, ditched the jargon, and translated it all into small biz speak. So you can actually useAI in your marketing… instead of Googling what “LLM” means for the fifth time.

 
 

🧠 Artificial Intelligence (AI)

Let’s start simple. AI is when computers do stuff that usually needs human smarts—like writing, learning, or making decisions.

Think of it like: a super-fast VA who never sleeps (but needs very specific instructions).

💬 Chatbot

That friendly (or sometimes too-friendly) little message box on a website that says, “Hi, how can I help?”

Biz use: Auto-replies to FAQs, bookings, or DMs so you don’t have to.

✍️ Prompt

What you type to get the AI to spit out something useful.

Tip: The better your prompt, the better your result.

Example: “Write a welcome email in a warm, funny tone that sounds like a yoga teacher and ends with a chai tea offer.”

✨ Prompt Engineering

A fancy name for learning how to ask AI the right questions to get what you want.

Your takeaway: You don’t need to be an engineer—just clear, specific, and a bit bossy.

📚 LLM (Large Language Model)

The brains behind AI like ChatGPT. It’s trained on heaps of text (books, blogs, the internet) so it can sound smart.

Think of it as: A well-read intern who doesn’t actually know anything—but can fake it brilliantly.

🧾 Tokens

The “currency” of AI. Every word or part of a word is a token. The longer your prompt or output, the more tokens it uses.

Biz use: Don’t overthink this unless you’re hitting limits on a paid tool.

🧰 Generative AI

AI that creates stuff—like text, images, music, etc.

Popular tools:

👯 “Act As” Prompt

Telling AI to play a role to get more targeted responses.

Example: “Act as a marketing coach and write a sales page for a life coach who works with burnt-out mums.”

Why it works: It gives the AI context—which means better content.

⚙️ Automation

When you connect tools to do things without you lifting a finger.

Example: When someone fills out your lead magnet form, AI auto-writes and sends a follow-up email.

Tools to try: Zapier, ManyChat, Mailchimp + ChatGPT integration.

📈 Machine Learning

The part of AI that “learns” from data over time. Not super relevant for everyday use—but it’s what powers your email inbox deciding what’s spam.

Your takeaway: Let the machines do their thing. You just need the results.

🧪 Training Data

What AI reads to learn how to sound smart.

Caution: It’s not always accurate or up to date (ChatGPT, for example, doesn’t always know the very latest info unless updated or connected to the web).

🚫 Hallucination

When AI confidently makes something up. Yes, really.

Example: “Here are 5 stats… that do not exist.”

Fix it: Always fact-check anything important.

📌 TL;DR

You don’t need to know all the AI jargon—just enough to use the tools well.

Want to skip the fluff? Our Prompt Packs do the heavy lifting for you, so you can just plug, paste, and post.