You've probably heard people talk about AI for the better part of two years. Maybe you've seen the LinkedIn posts. Maybe you've watched a colleague paste something into ChatGPT and get a surprisingly good answer. Maybe you've tried it yourself once, got a mediocre result, and quietly moved on.
That experience, trying it, not being impressed, moving on, is the most common entry point into this technology. And it's a shame, because the problem almost never is the tool. The problem is that no one tells you what question to actually ask.
This module fixes that. In the next 30 minutes, you'll produce one real artefact. Not a demo. Not a "fun experiment." Something you could actually use: a rewritten LinkedIn About section, a draft cover letter for a job you've seen, a study plan for a field you're thinking about moving into, or a sharp summary of a long article you've been meaning to read.
The tool we'll use is ChatGPT. Specifically the free version at chat.openai.com. It requires an account. It takes two minutes to sign up. You don't need to pay for anything to do this module.
One honest note before we start. This course will not teach you to build agents, write Python scripts, or fine-tune a model. It won't turn you into an AI engineer. It won't replace a proper careers counsellor, and it won't teach you the advanced operator-level workflows that people use professionally. Those exist, and some of them are genuinely remarkable, but they're not what beginners need on day one. What you need on day one is to finish something and feel the momentum. That's what this is for.
The first dumb question (and why it fails)
Most people open ChatGPT for the first time and type something like this:
"Help me write a cover letter."
The result is almost always usable and almost always disappointing. It's generic. It talks about "dynamic environments" and "leveraging skills" and hits all the right structural notes while saying essentially nothing about you. You paste it into a Word document, feel vaguely uneasy, and go back to the old version.
The problem isn't ChatGPT. The problem is that you gave it nothing to work with. You asked a vague question and got a vague answer. This is the number one thing beginners misunderstand about AI: it doesn't know who you are unless you tell it. It's not psychic. It's not reading your mind. It's trying to help a generic version of a person who hasn't told it anything.
The fix is simple. Give it context. The more context, the better the answer.
Here's what that looks like in practice.
The prompt that works
Pick your artefact. For this example, we'll use a LinkedIn About section rewrite. But at the end of the module there are versions for a cover letter, a study plan, and an article summary.
Copy this prompt. Fill in the bracketed parts. Paste it into ChatGPT.
---
Prompt, LinkedIn About section rewrite:
I need to rewrite my LinkedIn About section. Here's the context:
My name is [Name]. I currently work as [current role] at [type of company].
Before that I worked in [previous field or role, even if unrelated].
What I'm actually good at: [2 to 3 honest things you do well, could be communication,
staying calm under pressure, organising complex information, technical skills,
anything real].
What I'm trying to signal to the market: [What you want people to think when
they see your profile, are you positioning for a career change? Looking for
a new role in the same field? Trying to attract a certain type of client or employer?]
What I want to avoid: [Anything you hate about LinkedIn bios, corporate clichés,
overly formal tone, third person, anything else].
Write me three versions of my LinkedIn About section. Each should be between
150 to 250 words. First-person. Professional but not stiff. Australian spelling.
No exclamation marks. No phrases like "I'm passionate about" or "results-driven."---
Fill in every bracket before you paste it. Yes, even the "what I want to avoid" section. That part is more powerful than most people realise.
What you'll get back is three actual drafts, written in your voice as much as ChatGPT can approximate it, with specific constraints that filter out the junk.
The three edits that turn a draft into something you'd actually use
ChatGPT's first output is rarely the finished thing. It's the starting line. Here's how to close the gap.
Edit 1: Correct anything factually wrong. Read it carefully. AI makes things up sometimes, and in a biography that's a problem. If it adds a credential you don't have, a skill you didn't mention, or a claim that sounds off, fix it or tell ChatGPT to fix it. "That third paragraph makes it sound like I have management experience, which I don't. Rewrite it without that implication."
Edit 2: Inject one specific thing. Generic bios are forgettable. One real, specific thing makes a bio memorable. A project. A number. A city. A brief story. Ask ChatGPT: "Add one specific example or detail from my background that would make this less generic. Here's what you can use: [give it something real]."
Edit 3: Read it aloud. Seriously. Out loud. If you stumble over a sentence, that sentence needs to change. If it doesn't sound like how you actually talk, that's a problem. Go back to ChatGPT with: "This sentence sounds unnatural to me: [paste sentence]. Rewrite it in more conversational language."
Three rounds. Fifteen minutes. A LinkedIn About section you didn't have before.
What doesn't work
Pasting your old bio and saying "make this better." ChatGPT will polish what's there. If your old bio had the wrong positioning, you'll get a shinier version of the wrong thing. Start fresh with a context-first prompt.
Accepting the first output. It's never the best it can be. One round of edits almost always improves it. Two rounds usually nails it.
Asking for "professional" without defining what that means to you. Professional in construction sounds different from professional in finance. Tell it what your industry sounds like, or give it an example of writing you actually like.
What this means beyond the bio
You just did something that will matter more over the next three years than almost any other skill update you could make. You got a useful output from an AI tool by giving it context, reviewing the output critically, and editing toward your actual goal.
That process, in some form, is how every useful AI interaction works. The tool level up from here. The fundamentals don't.