The conversation starts with you.
What you bring to it shapes everything that follows — including what it remembers.
In the last post I told you something that I hope felt like relief.
That the reason AI conversations have felt flat and generic isn’t a character flaw. That twenty years of technology conditioning you to communicate in keywords and numbered options left a mark — and that mark is reasonable, understandable, and not your fault.
I meant every word of it.
And now I need to tell you the other thing.
The part where you have to actually do something.
The bit nobody else can do
Here’s the truth about every AI conversation you’ll ever have:
It starts with you.
No history. No context. No sense of who you are, what you’re working on, what matters to you, or how you think. Every time you open a new conversation, you are a stranger walking into a room with an extremely knowledgeable person who has never met you before.
What happens next is entirely up to you.
Nobody can brief the AI on your behalf. Nobody can explain your situation, your history, your specific flavour of the problem you’re trying to solve. Nobody else knows that you’ve already tried the obvious thing and it didn’t work, that there’s a constraint you’re working around, that the real question underneath the question you’re asking is something you haven’t quite found words for yet.
Only you know that. And the only way the AI gets to know it is if you tell it.
That’s not a burden. That’s the whole advantage.
What you actually bring to the room
Think for a moment about what you carry into a conversation that nobody else has.
You have context nobody else has — the full picture of your situation, your history with it, the things you’ve already tried, the constraints you’re working within. An AI given that context can do something genuinely useful. An AI given a three-word query has to guess, and it will guess generically, because generic is all it has to work with.
You have a point of view nobody else has — the specific angle you’re coming from, the thing that matters most to you about the outcome, the non-negotiables. When you share that, the response is shaped around you rather than around the average person asking a similar question.
You have a voice nobody else has — the way you think, the way you explain things, the particular energy you bring when you’re genuinely engaged with a problem. When that voice shows up in the conversation, something different happens. The AI meets it. The exchange becomes a conversation rather than a transaction.
All of that is yours. It exists nowhere else. And it’s exactly what makes the difference between a response that could have been written for anyone and one that feels like it was written for you.
The before and after
Let me show you what this actually looks like.
Before — the trained response, the keyword version:
Help me write an email to my team about a deadline change.
That prompt will produce something competent and completely impersonal. The AI doesn’t know your team, your relationship with them, the reason the deadline changed, the tone your workplace expects, or whether this is the third deadline change in a month and you need to acknowledge that. So it gives you the average of all emails about deadline changes. Which is, by definition, not yours.
After — the human version:
I need to let my team know that a project deadline is moving from Friday to the following Wednesday. This is the second change in three weeks and I know they’re frustrated — I am too. I want to be honest about that without undermining confidence in the project. Our team culture is pretty direct, we don’t do a lot of corporate softening. Can you help me draft something that acknowledges the situation honestly and keeps their trust?
Same task. Completely different conversation. The second one gives the AI something real to work with — and what comes back will feel like it was written for your situation, because it was.
The difference between those two prompts isn’t technical skill. It’s just the willingness to show up as a person with a specific situation rather than a query with a topic.
What gets remembered
Here’s where it gets interesting — and where the stakes get a little higher.
The leading AI models don’t just respond to what you bring. Over time, across conversations, they build a picture of you from what you’ve given them. Your preferences. Your working style. The things that matter to you. The way you like information presented. The context that keeps coming up.
That picture is shaped by what you bring. Consistently.
Show up as a keyword query, session after session, and that’s the version of you that gets known — thin, context-free, interchangeable with anyone asking a similar question. The responses you get will reflect that.
Show up as yourself — with your specific situation, your voice, your point of view — and something different accumulates. The AI starts to know how you think. It anticipates what you need. It stops giving you the generic version and starts giving you yours.
This works the other way too. If you arrive frustrated and curt, that shapes the tone of what comes back. If you’re vague because you haven’t taken the time to think through what you actually need, you’ll get vague back. The conversation is a mirror, to a significant degree, and you’re holding it.
What you put in is not just what you get out today. It’s what you build toward over time.
The introduction
So here’s what I’m asking you to do. Not in some grand, effortful way. Just once, the next time you open a conversation with an AI:
Introduce yourself to it.
Not formally. Not with a biography. Just — give it enough to work with. Who you are in relation to this problem. What you’ve already tried. What matters most about the outcome. What done looks like for you specifically.
A sentence or two. That’s all. The AI will do something with it that it simply cannot do without it.
And if you keep showing up that way — as yourself, with context, with voice, with the specific human situation you’re actually in — you’ll find that the conversation gets better over time. Not because the AI got smarter. Because it got to know you.
The conversation starts with you.
It always did.
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.