HumanPrint: How to Use AI Without Losing Yourself
Jul 23, 2026 · 13 min · 10 segments
**What this episode is about** Every time another "how to spot AI writing" post shows up, my response is the same: how do I know YOU wrote it? Before you accuse anyone else of using Gen AI, can you…
Christine WhitmarshHostToday's episode is called, How Do You Sound Like You? You know, every time I see one of those social media posts, you know, here's how to know if AI wrote it, my response, usually internally, is the same.
How do I know if you wrote it? Before you start accusing everyone else of using AI, can you actually identify what makes your own writing recognizably yours? So yeah, welcome to Know Thyself, Writing Edition.
They collect a list of phrases they associate with AI and they start cutting them out.
Certain transitions, punctuation, sentence patterns, whatever the AI gotcha police have decided this week is proof of bot activity.
And even if they do catch every supposed tell, which they probably won't, they're left holding a more important question.
What remains? Removing generic language doesn't automatically reveal your voice.
Spotting AI writing, it's getting easier and easier, especially when someone's pasting in raw, unedited output.
These systems are trained on enormous amounts of human language, and they generate text by modeling statistical relationships within that data.
When they have no real information about you, well, they're just going to generate random things from that big generic mixture.
That random mixture of words, that's not automatically your identity just because you attached your name to it.
Back in the early chatbot days a few years ago when I asked a generic LLM to write something for me as an experiment, I was the part that disappeared, me.
The writing, it might have been clean and grammatically correct and reasonably well-structured, but it just didn't sound like me because the system had almost nothing to work from.
Before I built my own voice intelligence system, custom bots trained around a person's actual writing voice and point of view, the tool was just pulling from general patterns.
It needed to be working from me, not from just a big data set, which meant I had to understand what I sounded like on the page.
Fortunately, I've been writing forever and I've been ghostwriting books since 2001, which means I've spent decades studying how other people think, speak, tell stories, explain ideas, argue, joke, relate to a reader, and move through sentences.
At its best, ghostwriting is not collecting someone's favorite words and sprinkling them into a manuscript.
There's words and grammar, but there's also rhythm, sentence length, layering, directness, emotional range, how people use examples, humor, questions, metaphor, story, certainty, contradiction, pacing, how the writer positions themselves.
When I zoom out in my own writing, I see the curious scientist, the lifelong creative, the storyteller, the writing coach, the habits enthusiast who keeps wandering into psychology, someone who resists black and white conversations and loves language just enough to play with it all the time.
Today's episode is called, How Do You Sound Like You? You know, every time I see one of those social media posts, you know, here's how to know if AI wrote it, my response, usually internally, is the same.
How do I know if you wrote it? Before you start accusing everyone else of using AI, can you actually identify what makes your own writing recognizably yours? So yeah, welcome to Know Thyself, Writing Edition.
They collect a list of phrases they associate with AI and they start cutting them out.
Certain transitions, punctuation, sentence patterns, whatever the AI gotcha police have decided this week is proof of bot activity.
And even if they do catch every supposed tell, which they probably won't, they're left holding a more important question.
What remains? Removing generic language doesn't automatically reveal your voice.
Spotting AI writing, it's getting easier and easier, especially when someone's pasting in raw, unedited output.
These systems are trained on enormous amounts of human language, and they generate text by modeling statistical relationships within that data.
When they have no real information about you, well, they're just going to generate random things from that big generic mixture.
That random mixture of words, that's not automatically your identity just because you attached your name to it.
Back in the early chatbot days a few years ago when I asked a generic LLM to write something for me as an experiment, I was the part that disappeared, me.
The writing, it might have been clean and grammatically correct and reasonably well-structured, but it just didn't sound like me because the system had almost nothing to work from.
Before I built my own voice intelligence system, custom bots trained around a person's actual writing voice and point of view, the tool was just pulling from general patterns.
It needed to be working from me, not from just a big data set, which meant I had to understand what I sounded like on the page.
Fortunately, I've been writing forever and I've been ghostwriting books since 2001, which means I've spent decades studying how other people think, speak, tell stories, explain ideas, argue, joke, relate to a reader, and move through sentences.
At its best, ghostwriting is not collecting someone's favorite words and sprinkling them into a manuscript.
There's words and grammar, but there's also rhythm, sentence length, layering, directness, emotional range, how people use examples, humor, questions, metaphor, story, certainty, contradiction, pacing, how the writer positions themselves.
When I zoom out in my own writing, I see the curious scientist, the lifelong creative, the storyteller, the writing coach, the habits enthusiast who keeps wandering into psychology, someone who resists black and white conversations and loves language just enough to play with it all the time.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
