What is ownership?
When AI writes the words, ownership lives in the decisions, not the keystrokes.
A few weeks ago I was explaining to our content marketeer how I write these blog posts. I walked her through the whole process. How I come up with the topics. How I find the angle, the hook. How I figure out what’s actually interesting to say versus what’s just noise. How I pace the thing, where the tension goes, what the reader should feel at the end.
All me. Every bit of that is me.
Then I told her I have Claude write the first draft.
Her face changed. Not dramatically. Just a slight shift, like when someone recalculates something mid-conversation. She’d been nodding along, following the creative process, seeing the authorship. And then, in one sentence, it moved. To her, the moment I said “Claude writes the draft,” the steering wheel changed hands.
To me, nothing had changed at all.
The line people draw
She wasn’t wrong to react that way. If you think of writing as the act of putting words on a page, then the ‘person’ putting words on the page was Claude. That’s just true. The sentences, the paragraph breaks, the word choices in that first draft: Claude.
But if you think of writing as deciding what to say, then none of that matters. The topic was mine. The angle was mine. The hook was mine. The structure, the examples, the things I specifically wanted to include or avoid, the voice, the perspective, the opinion. All mine. Claude got a rambling mess of thoughts and turned it into paragraphs. Which I then further refined and rewrote. Together with Claude.
Is that different from talking to a ghostwriter? From dictating into a recorder and having someone transcribe and clean it up? From sketching on a napkin and having someone render it in Figma?
The content marketeer saw a clear line. On one side: you improving your own draft with AI. That’s fine. On the other: AI generating text from your input. That’s different. That’s not you anymore.
I don’t see that line. Or rather, I see it, but it’s in a completely different place than she thinks.
Where the actual steering happens
Ownership lives in the decisions, not in the keystrokes. Did I decide the topic? Yes. Did I decide the audience? Yes. Did I decide what’s interesting about this and what’s boring? Yes. Did I read the draft and think “no, that’s not what I mean” and redirect? Multiple times.
The draft is a starting point. It’s clay. Sometimes the clay is pretty close to what I wanted. Sometimes I reshape most of it. Sometimes I keep a sentence that’s better than what I would’ve written and think “yeah, that’s exactly it, I just wouldn’t have phrased it that way.”
Is that sentence still mine?
I think so. I chose to keep it. I recognized it as right. That recognition is a creative act.
A photographer doesn’t manufacture the light. They recognize the moment and press the shutter. Nobody questions their ownership of the photo.
The same thing happened with code
This isn’t just a writing thing. I’ve watched my own relationship with code follow the exact same arc.
Two years ago, I was dictating code. Line by line, sometimes. “Create a function that takes these parameters, validates this, calls that API, handles these errors.” The AI was a fast typist. I was still the programmer.
Then I moved up a level. I stopped dictating code and started dictating architecture. “Build a service that handles webhook ingestion. It should queue events, deduplicate based on this key, and process them async with retry logic.” I wasn’t writing functions anymore. I was designing systems and having the functions written for me.
Then another level. I stopped dictating architecture and started discussing it. “I need to handle webhooks from this provider. They’re unreliable and sometimes duplicate. What’s a good approach?” A conversation. Back and forth. The AI suggests, I push back, it adjusts, I refine.
And now? I’m discussing the feature. Not the code, not the architecture. “Users are confused about the status of their request. They keep contacting support to ask what’s happening. What should we build?” That’s where the conversation starts now. The code, the architecture, the implementation, those fall out of it.
Each transition felt like giving something up. Each one was actually moving up.
Then what is ownership?
The content marketeer’s discomfort wasn’t irrational. There’s a real question buried in it.
If the value of a writer is in the words, and the words are generated, what’s left? If the value of a programmer is in the code, and the code is generated, what’s left?
The answer, I think, is that the value was never in the words or the code. It was in knowing which words and which code. It was in taste. In judgment. In having something to say in the first place.
A blog post nobody wants to read is worthless regardless of who typed it. A feature nobody needs is worthless regardless of how clean the implementation is. The typing was never the hard part. It just felt like it was, because it took the most time.
I still own these blog posts. Not because I typed every word. Because every word is here because I decided it should be.
The line keeps moving
The interesting thing about my content marketeer’s reaction is that her line will move too. It already has, probably, since that conversation. The first time she uses AI to draft something and thinks “this is exactly what I meant, it just came out faster,” the line moves. Not because she was wrong before. Because the tools redefined what “doing it yourself” means.
Two years ago, writing code meant typing code. Now it means describing what the code should do and why. Two years from now, who knows. The line between “my work” and “generated for me” will keep shifting, and at every step, someone will draw it exactly where the previous generation stopped being comfortable.
The question isn’t whether AI wrote this. The question is whether this says what I wanted to say, in the way I wanted to say it.
It does. So it’s mine.
Figuring out what’s mine at neople.io.