It’s Not You. It’s the Question.
Why AI keeps letting you down, and what nobody bothered to tell you
You tried it. Of course you did.
Everyone’s been talking about it — at work, at dinner, on the news, in your social media feed at eleven o’clock at night when you should be asleep. So you opened one of those apps, typed something in, and waited.
And what came back was... fine. Generic. Vaguely helpful in the way that a fortune cookie is vaguely helpful. Nothing that made you think “oh, THIS is what everyone’s been going on about.”
So you tried once more, got something similar, and quietly closed the tab. And somewhere in the back of your mind, a small voice said: “I must be doing it wrong. This probably isn’t for people like me.”
I need you to hear this clearly:
That voice was wrong.
The problem wasn’t you. It wasn’t your intelligence, your age, your technical ability, or your right to be in the room. The problem was that nobody told you the one thing you actually needed to know before you started.
AI doesn’t respond to keywords. It responds to context.
We’ve spent twenty years training ourselves to talk to search engines — short, clipped, efficient. Three words and a question mark. ”Best Italian Melbourne.” “Symptoms fever headache.” “How fix leaking tap.” That’s the language of Google, and Google understands it perfectly.
But AI isn’t a search engine. It’s a conversation. And conversations need more than three words to go anywhere useful.
When you type “write me an email” and get something stiff and generic back, that’s not the AI failing you. That’s the AI doing exactly what you asked — with almost no information to work with. It doesn’t know who you’re writing to, what the relationship is, what tone you want, what outcome you need, or what you’ve already tried. So it gives you the average of everything. Which is, by definition, nothing special.
Here’s what that looks like in practice.
The prompt that gets slop:
Write me an email to my boss about taking time off.
The prompt that gets something useful:
I need to ask my manager for two weeks off in July for a family holiday. We’re a small team and it’s a busy period, so I want to acknowledge that upfront and offer to prepare a handover plan. My manager is practical and direct — she doesn’t need a lot of softening. Can you help me draft something professional but warm?
Same task. Completely different conversation. The second one gives the AI something to actually work with — and what comes back will feel like it was written for you, because it was.
That’s not a technical skill. That’s just... talking. The way you’d talk to a colleague you trust.
Now let’s talk about the other fear. The one that’s harder to say out loud.
What if AI replaces me?
I hear this everywhere — from friends, from people I’ve worked with, from anyone who’s built a career on doing something well and is now watching a machine apparently do it in seconds. The fear is real and it deserves to be taken seriously, not dismissed with cheerful reassurances about how technology always creates new jobs.
But here’s what I’ve observed, and I want you to sit with this for a moment:
The people most at risk from AI are not the ones who can’t use it. They’re the ones who refuse to.
Because the gap between someone who knows how to have a good conversation with AI and someone who doesn’t — that gap is widening every single day. And it’s not a gap that requires a degree, or a course, or a technical background. It requires exactly what you already have: the ability to explain what you need in plain language, with enough context for someone else to actually help you.
You’ve been doing that your whole life. You just haven’t been doing it with this particular listener yet.
I built an app for this.
Not because I’m a developer — I’m not, at least not in the traditional sense. But because I got so frustrated watching intelligent, capable people give up on AI simply because nobody had shown them how to have a decent conversation with it, that I sat down and built something to help.
Here’s what it does: before you type a single word into your AI chatbot, the app asks you the questions. What are you trying to do? Who is it for? What tone do you need? What matters most? In just a few clicks, it assembles all of that into what the tech world calls a prompt — a properly formed question that gives the AI everything it needs to give you something genuinely useful back.
You don’t need to know what a prompt is. You don’t need to learn any new language. You just answer a few simple questions, copy what the app gives you, and paste it in.
That’s it. No technical knowledge required.
I want to show you something.
I asked an AI to generate an image of a man pulling on a push door. It took six attempts and the door was open every time.
Which, now that I think about it, is rather the point.
The door was always there. The handle was always there. He just needed someone to tell him which way it opened.
The door opens from this side.
Not from the search bar. Not from three words and a question mark. From the human side — the one with context, and history, and a real situation you actually need help with.
Because that’s the whole thing, really. Starting the conversation. Giving it enough to work with. Treating it less like a search engine and more like a very patient, very knowledgeable colleague who genuinely has time for you.
You’re not behind. You’re not too old. You’re not too non-technical.
You just needed someone to tell you that the door opens from this side.
Sandi is a Melbourne-based problem-solver, crisis-averter, and translator of the technical into the human. She spent decades being the person everyone called when something was broken, confusing, or just needed explaining properly — earning a reputation that preceded her wherever she went. Now she’s channelling that same instinct into AI: making it accessible, practical, and genuinely useful for people who think it isn’t for them.