The Question Everyone Asks
People keep asking me the same question.
“Is this you or is this AI?”
I get it. The last 20–30 posts look cleaner. The structure is tighter. The English is better. I’m not a native speaker and I never pretended to be. So the suspicion is fair.
My Actual Writing Process
Here’s the actual process.
I open Wispr Flow. I talk. Most of the time it feels like a short interview with myself. I dump the raw thoughts while walking, while driving, while waiting for coffee. No outline. No carefully chosen words. Just the thing that is actually sitting in my head that day: something that worries me, something I’m testing, something that feels urgent in the work I’m doing.
Then I take that transcript and I use AI the same way I use any other leverage tool. I clean the syntax. I fix the grammar. I tighten the sentences so the idea doesn’t get lost in the noise of my imperfect English. That’s it.
So Is It AI-Written?
Is the post AI-written? No. The core is still the same messy, personal material it has always been. The only difference is that I no longer spend three or four hours fighting the language. I spend thirty minutes. And the output is usually better because the energy goes into the thinking instead of the polishing.
I used to believe that “real” writing required suffering through every sentence. That belief was mostly ego. Time is the scarcest resource I have. I’m a solo operator building systems that are supposed to multiply intelligence, not consume it. If a tool lets me protect the signal and discard the friction, I use the tool. Full stop.
What I Refuse to Outsource
What I refuse to outsource is the source material itself. The topics are still the things I actually care about day to day. The doubts about agent reliability. The tension between speed and safety in multi-agent systems. The quiet realization that most “AI strategy” is still theater. These are not generated ideas. They are the ones that keep me up or push me forward. AI doesn’t invent them. It just stops me from butchering them on the way out.
Authenticity Was Never About the Prose
There’s a deeper point here that most people miss when they obsess over the “human vs AI” binary. The valuable part was never the perfect prose. It was the lived experience behind the prose. People tell me they find value in the posts because they recognize the texture of real decisions, real trade-offs, real constraints. That texture doesn’t come from a model. It comes from the person who is actually in the arena.
I could spend hours making every sentence sound more “authentic.” Or I can ship the thought while it’s still sharp and move on to the next experiment. I choose the second option. Deliberately.
Some will still say the posts feel too polished now. That’s fine. Others will claim that using any AI assistance makes the whole thing inauthentic. Also fine. I’m not trying to win a purity contest. I’m trying to transfer useful signal as efficiently as possible while I still have the energy and the context to generate it.
The irony is that the more I lean on these tools, the more personal the content becomes. Because the bottleneck is no longer language. The bottleneck is honesty about what I’m actually seeing and testing. And that bottleneck is still 100% human.
So yes, AI is in the loop. Wispr Flow is in the loop. But the questions, the worries, the observations, and the conclusions are still mine. If that bothers you, you are free to stop reading. If it doesn’t, then we can keep going.
I’m not interested in performing authenticity. I’m interested in staying useful.
Join the Discussion on X
This piece is also running as a discussion on X — reply there or in the comments below.
Leave a Reply