What AI actually is

The 60-second version

ChatGPT, Claude, Gemini, and Copilot all run on the same kind of engine: a large language model, or LLM. It has read a huge amount of text and learned the patterns in it.

When you type something, it does one thing: it predicts the most likely next words. Over and over, very fast.

That sounds too simple to be useful. It isn’t. Predicting language well means the model absorbed how contracts read, how apologies sound, how a job post is built. So you get a tool that can read, write, summarize, and explain — instantly, about almost anything.

Fluent — and sometimes fluently wrong

The model produces plausibletext. It doesn’t look up verified facts. So it can state something false with total confidence — an invented date, price, or rule. That’s called a hallucination.

The good news: the failure is narrow. The writing is usually solid. It’s the specific facts you check. Module 5 turns that into a one-minute habit.

What it's great at — and bad at

The last one matters most. The model knows nothing about your business until you tell it. Telling it well is the whole skill — and it’s Module 2.

Which tool should you use?

Any of them. All four have free tiers that handle everything in this course. If you already pay for Microsoft 365 or Google Workspace, you may have one included.

Three words you'll hear constantly

LLM

The engine under these tools. Trained on huge amounts of text to predict what comes next.

Prompt

Whatever you type to it — instructions, questions, and anything you paste in.

Hallucination

A confident, made-up detail. The reason you verify facts.