What ChatGPT is, and how to talk to it.

If you've heard everyone talking about ChatGPT and quietly wondered what the fuss is, this is for you. Five short lessons, plain English, no code. By the end you'll be using it for real work. Let's start with what it actually is, and the one habit that makes it genuinely useful.

The brilliant brain you chat with

The simplest way to picture ChatGPT is as a brilliant brain you have a conversation with. It has read an extraordinary amount, and it can write, explain, summarise, translate and reword in seconds. You type a request in plain English, the way you'd ask a sharp new assistant, and it writes back. That's the whole interaction. No buttons to learn, no special language.

One honest caveat, and it matters from day one: a brilliant brain can still be confidently wrong. It will sometimes state something untrue as though it's certain, especially with names, numbers and specifics. So treat it as a quick, capable first-drafter, not an oracle. For anything that counts, you check the work. We give that its own lesson later, because it's the difference between a tool you trust and a mess you clean up.

It also helps to know what it isn't. It isn't a search engine reaching out to live web pages, and it isn't plugged into your business: it doesn't know your prices, your customers or last week's emails unless you tell it. Picture a well-read person who's just walked in the door. Sharp, but they only know what you put in front of them.

Free or paid: which do you need?

You can do a great deal on the free version, and it's the right place to start. The paid plan (and the Team plan for businesses) adds the smarter models, fewer limits when you're busy, and the ability to build Custom GPTs, which we cover in lesson three. A fair rule of thumb: learn on the free tier, and only pay once you've found a job you'd happily use every week. There's no need to spend a cent to get through this course or to get value this week.

How to talk to it: the recipe that works

This is the part that separates a vague, generic reply from one you can almost use as is. The trick is to stop typing one-line questions and instead set the scene. A reliable recipe, in five quick parts:

• Give it context: who you are, who it's for, what you're trying to achieve. "I run a two-van plumbing business in Geelong."
• Give it a role: tell it the hat to wear. "Act as a friendly customer-service writer for a trade business."
• Show an example: paste a past email or note that's in your style, so it copies your tone rather than inventing one.
• Say the format you want: "Keep it to four short sentences, warm but professional, no jargon."
• Then refine: read the first answer and tell it what to change. "Shorter, and mention we can come Thursday." It adjusts on the spot.

You won't always need all five. For a quick reword, "make this clearer and a touch warmer" is plenty. But when an answer comes back flat or generic, it's almost always because the request was thin. Add context and a role and watch it lift.

Before and after

Here's the same job done two ways. The weak prompt: "Write an email to a customer about a late delivery." You'll get something stiff and generic that could be from anyone.

The strong prompt: "You're writing on behalf of a small family furniture shop. A regular customer's order is running a week late from the supplier. Write a short, genuine apology email: warm, plain English, no corporate waffle. Offer free delivery to make up for it, and give a new date of next Friday. Four sentences." Now you get a draft that sounds like you and is ready for a quick read-over. Same brain. The only thing that changed is what you gave it.