Tom Lambotte / Writing with AI
Yes. Most readers spot AI writing within a few seconds, and they rarely tell you. They disengage.
Tom Lambotte has coached 225 founders through this. Here is how to keep your voice while you use AI every day.
Yes. When I ask a room of founders if they can spot AI-written content, every hand goes up. They can all spot it within a few seconds.
I got an email from someone I've trusted for years. Two sentences in, I knew she didn't write it. The idea was hers, but every word screamed written by AI.
I stopped reading. I didn't unsubscribe, reply or mention anything. I just disengaged.
That is what most readers do. A 2025 study by Raptive (n=3,000 U.S. adults) showed 52% of readers disengage the moment they detect AI-written content.
It costs you trust, and you won't see it leaving. When your audience spots AI-generated content from you, they don't write to tell you how frustrated they are. Nor do they unsubscribe. They disengage.
Imagine an email list of 20,000 people you built over 10 years. Every email that reads like AI damages your trust. The list size doesn't move, nor do your open rates shift. But people who used to forward your emails no longer do. Clients who chimed in go quiet. Prospects you were warming up go cold.
You won't notice anything over 30 days. Within six months, your conversion rate is down. You're blaming the offer. In 18 months, you're wondering what changed. The damage is invisible until it is structural.
It happens inside your own company too. I asked a founder on a workshop if he had stopped reading anyone because he could tell AI wrote it. He said yes: his own team. He owns a group of real estate companies.
"I used to read every email from my companies. Now that they are using AI, I just delete them. I don't want to read them anymore because it's written by AI. I used to be interested in what they were coming up with. I'm not interested in what their AI bot is coming up with."
Because your voice gets altered in the last 10 to 20%, the part you let AI finish. Most founders I see do the first part well. They feed AI their original thinking, source content, and unique worldviews.
I read one author's newsletter for ten years and finally stopped. I told him his AI was showing. He got defensive and said it is all his own material, his team only uses AI for the last 10 percent, to finish it up.
That last 10 percent is where his unique, authentic voice lived.
AI does its best work in the first 90 percent: the thinking, research, structure, getting unstuck on a blank page. It does the most damage in the last 10, because the closing line, the odd word choice, the aside nobody else would make, that is the part your reader recognizes as you. Hand that piece over and you sound like everyone else.
Even she hit the wall so many have: AI content that sounded flat, fake, and painfully cliché.
The solution was learning the specific, nuanced way to get AI to capture her rhythm: the tone and emotion that make her unmistakable.
"Once AI finally started sounding like me, the shift was instant," she told me. "Our clients noticed, and my team felt the difference."
The two patterns I trip over most are the contrast flip and rhythmic stacking.
The contrast flip is the "not this, but that" shape, as in "Not a strategy problem, a clarity problem." Readers have scrolled past thousands of AI posts built on that exact shape, and they feel it before they consciously spot it.
Rhythmic stacking is ideas listed as separate one-word sentences in parallel, as in "Different expert. Different framework." It reads as generated because it usually is.
Yes, and it can take twenty minutes. A founder I worked with had just spent $8,000 on a brand new website. The agency he hired had quietly used AI to write the words, and they had done a sloppy job of it. His homepage, the front door of a luxury travel business built on decades of relationships, was speaking in a voice that was not his. When we ran the AI detection scan together, it scored 78.
In twenty minutes we put it back. We cleaned four patterns out of the copy:
What replaced it was his actual language, the way he sounds when a longtime client calls on a Tuesday afternoon. Score after rewrite: 5 out of 100, where lower is better. He got his voice back.
Talk your idea out first, then make AI keep your words. When AI's draft starts sanding down your edges, removing quirks, smoothing over your phrases, push back. Tell it: "No, use my original words. Keep my tone. Keep my examples."
The step-by-step is on how founders talk to AI instead of typing.
Yes. We built one. It's called AuthentiVoice, and everyone in The 4-Week AI Leverage Sprint for Founders gets access to it.
A lot of the people who can spot AI writing take a moral stance: "I will not use AI. It sounds fake." I get it.
But what if you could use AI in a way that maintains your AuthentiVoice? Where Susie sounds like Susie and communication from your company sounds like communication from your company?
Authenticity is going to be the most valuable currency in the coming age.
Pick one thing going out under your name this week. Read it out loud before it goes. If it does not sound like you talking, your reader will notice before you do.
Then go into your sent folder, or the sent folder of your leadership team, and read the last five emails they sent. Ask yourself, does this sound like a human wrote it or AI did?
If you lead a law firm and wonder whether using AI every day adds up to fluency, start with this page. Keeping your voice while you use AI is the work we do together in The 4-Week AI Leverage Sprint for Founders.
The 4-Week AI Leverage Sprint for Founders. Four live sessions. $5,000.
See the Sprint