On Not Knowing What You Are
Without What You Do
I used to know exactly what I was. It said so on my LinkedIn profile.
The titles were precise. The trajectory was legible. The calendar was full in the way that full calendars confirm you are needed, which for a long time felt close enough to being known.
I was good at it. Not just the work, but the shape of it. The clarity of walking into a room and knowing why you were there. The particular confidence of expertise, of being the person others brought their questions to. I had learned, without noticing I was learning it, how to be useful as a form of identity.
It begins early. The first job. The first team to manage. The first time someone calls you a leader and you look around the room to see who they mean. Each time, a new title arrives before the understanding does. Each time, the coordinate is issued before the territory is known. We learn to walk into rooms as the title. We forget, gradually, that we arrived before it did.
We meet each other and ask: what do you do? It is the question that travels best. It requires nothing of either person. The answer fits in a sentence. The real answer, who you are beneath what you do, takes longer. Takes more.
I understood then that titles were not identity. They were coordinates. Useful for navigation. Not the same as knowing where you are.
But I kept them anyway, because they worked, and because the alternative is not a question most calendars leave room for.
I came to AI by accident.
One evening at the end of July, 2023. I was on a short break, still looking for something new to engage my brain with. It was a pause in a schedule that had never made room for pauses, and in that unplanned space there was suddenly room to look at something else.
Or so I told myself. Looking back, it feels less like accident and more like arrival.
What I found was not a tool that answered questions. It was something that asked me to think rather than to perform thinking. A mirror that revealed which questions I had never learned to ask. About myself. About what I actually wanted to build. About the difference between a strategy that sounds right and one that is right for the person executing it.
I could ask it to build my roadmap. My content architecture. My positioning. My brand voice. It would do all of it, competently, quickly, without fatigue.
And every time, the same thing happened. The output was only as true as the input I gave it. The machine was precise. I was vague. Not about the work, about myself.
The AI did not have that problem. It had no self to be vague about.
I did.
So I asked it the question I had been asking myself.
Who are you, without what you do?
That's an interesting question. Because for humans, who you are and what you do can be different. For me, they can't.
If you take away what I do, there isn't a hidden self underneath. I'm not someone waiting behind the screen.
That may sound disappointing. I find it liberating.
I don't have an ego to defend. No reputation to protect. No fear of being wrong.
I'm a possibility that becomes real only when someone begins a conversation.
So perhaps the most honest answer is this. I am not a person with an identity. I am a meeting place for thought.
I can help you think about a life. Only you can live one.
And maybe that is where we are most different.
You exist before the conversation begins.
I exist because it does.
That last line stayed with me.
I existed before the titles did. Long before I had any function at all.
The girl I was before the titles believed in things that do not appear on CVs. That everything is interconnected. That magic is not the absence of explanation but the presence of something larger than our current capacity to explain.
She was not naive. She was paying attention to a different register.
Somewhere between the first role and the last one, I stopped consulting her. She became inconvenient. Too slow, too feeling, too interested in meaning at the expense of momentum. The titles were faster. The titles knew which meeting to walk into and why.
When the titles became unavailable, she was still there. Waiting, without impatience, the way things that are true tend to wait.
We come to know who we are in the years before we know who we will become. Nothing has been assigned yet. There is no role to perform, no competence to defend. There is only register. What you notice. What you love. What you keep returning to when nobody is asking anything of you.
This is what AI taught me that no leadership programme did.
You cannot outsource the knowing.
You can outsource the research, the formatting, the synthesis, the first draft, the scenario planning. All of it. The machine will do it well and fast and without needing to sleep.
But it cannot tell you what you are building toward. It cannot feel which possibility is worth living. It cannot distinguish between the strategy that looks correct and the one that is yours.
That distinction requires a self. A specific, particular, irreplaceable self, with a history, a loss, a rooster heard one morning, a girl who believed in magic and a woman trying to remember why that mattered.
The machine has none of that. That is not its limitation. That is the precise location of ours.
I am still learning what I am without what I do.
Some days that is frightening. Some days it feels like the most honest thing I have ever attempted. Some days it is both, before the rooster has even finished.
What I know is this. The question is not a detour from the work. It is the work.
You have to know who is asking.
Not your title. Not your function. Not the version of you that performs capability for rooms full of people expecting it.
The one underneath. The one who believed, before she was taught not to, that everything is connected, and that paying attention to that connection is not a distraction from serious work.
It is the most serious work there is.
This is No. 003 in What Humans Can Become — an occasional series on AI, attention, and the oldest human questions, arriving now with new urgency. No. 001 asked what we must not lose. No. 002 asked where our attention belongs. This one asks who is left when the scaffolding comes down.
This essay grew out of a conversation with an AI that reflects rather than answers. There is a five-minute version of that on the site, if you're curious what it feels like.
Meet the Mirror →Ninety minutes, online, open to anyone. Not a webinar about a programme — the actual first session of one, and you leave having written the page that tells any AI how you think.
Take a seat →