It was not the question I expected to come home with.
I recently pushed myself out of my comfort zone and went on my first-ever walking weekend in Wales with 35 women. I knew a handful of them, but not overly well. Fortunately, within minutes of arriving, I felt truly welcome.
One moment in particular stayed with me. We were standing around the large kitchen island, everyone chatting, eating their morning porridge, or handing out coffee.
And AI came up.
Not in a tech room. Not on a business webinar. Around a kitchen island in Wales, with women I had only just met.
It reminded me how far this has already spread. Whether you use AI or not, it is probably touching your working day in some way.
I do not think there are only two camps: people completely for AI and people completely against it. There are also people like me, trying to use it carefully, without letting it flatten how we actually think and help people.
Before you read on, never think I am saying this because I have mastered it. I have not.
I am saying it because I went too far.
Twelve months ago, I was walking a very thin line with AI. I had lowered my standards of what was okay, handed over too much of my thinking, and ended up jumping to AI for any question I had, taking the answer as gospel without first switching my brain on.
At first, I knew what I wanted to achieve and was guiding the tool properly. But because AI is so fast, you can get caught up in that speed and start skipping steps you know you should be doing.
The draft looks structured. It reads fine. There is nothing obviously wrong with it, so you skim it, accept it, and cross it off your to-do list because AI has given you something polished.
But there is a skill to using AI.
Take it from me, you cannot just jump to the tool and let it do the work for you. If you do, you end up editing manually, going back to the tool again and again, or posting content that does not reflect your true expertise. I should know. I have done all three.
And there is zero point in sharing content that is slowly damaging your credibility. Zero point.
Most people I speak with do not want to stop using AI. They want the speed, but they do not want to sound bland or AI-ish. They want to carry the experience people already trust them for.
That is where clarity before AI matters.
AI will fill in the gaps if you have not done the thinking first. But it cannot fill them with your lived experience, your way of seeing a problem, or the small details that make your work feel like yours. It will fill them with average words.
One conversation in Wales was about someone who had passed off an AI critique as his own. She could tell it was AI, and his credibility dropped quickly. Another person passed off work as his own, but when asked a clarifying question, he fell apart because he had not read or digested the output.
In an office, that kind of thing can become visible quickly. People notice. People talk.
Online, I think it is different. Your audience might not say, “That was definitely AI.” They might just slowly engage less, reply less, click less, enquire less, and take less notice.
The bigger risk is not being caught using AI. It is whether you start becoming easier to ignore, while someone else becomes more present in your audience’s mind.
And you might be thinking, “But I need AI because I do not have time to create content properly.”
I understand that. But if the content you create faster does not show the way you think, the way you solve problems, or the standard you would bring if someone was sat in front of you, then the speed win is not really a win.
Maybe you are also thinking your audience will not notice. Maybe not directly. But if your content starts to feel like everyone else’s, people do not have to call it out for it to matter. They can simply stop caring as much and move on. Ouch.
I also would not recommend just editing your AI draft to make it sound more like you. I think that is too late. I know because I did it. I would remove phrases I would not say, change words, rip out sentences, and ask the tool for another rewrite. Any time I thought I had saved was gone.
The problem was not only the wording. The thinking was not strong enough at the start. I had become too focused on the perfect structure, the perfect hook, and the “award-winning” version of the post. I was not focused enough on what needed to go into it: my lived experience, opinions, personality, and actual presence.
My time on LinkedIn exposed that gut feeling. I could feel when other people showed how they thought and solved a problem, and I sensed I was not creating that same presence in my own content.
So I started pulling back. It never crossed my mind to stop using AI; it is not the villain in all of this. I just became obsessed with understanding how to use it.
Prompts help. Business context helps. Strong instructions help. But a prompt does not capture your thinking on its own. It does not automatically know what you believe, what you would challenge, or the standard you bring when working. That has to come from you.
And yes, that can feel slower at first because we are promised speed, scale, and more content with less effort. But before you get the real speed win, you may need to slow down enough to move your effort upstream.
You do not want AI amplifying weak thinking. You want it to amplify clarity and full thinking. When you do that, there is less to fix afterwards, fewer rewrites, less second-guessing, and more confidence that the content reflects the work people already come to you for.
This is why I care so much about what happens before AI writes.
The strongest content does not start with a prompt. It starts with clarity. It starts with capturing your raw thinking, pulling out the useful stuff, and strengthening it first so AI has something real to work with.
If that is the part you know you need help with, you can book a 30-minute fit call to see if the voice-to-content sprint is right for you.
The goal is not to use AI less. It is to use it without handing over the thinking and presence that made people trust you in the first place.
Keep thinking clearly,
Shannon
