Why AI Doesn't Sound Like You (and the One-Line Fix)
Updated: Jul 31
You can usually tell when something was written by AI. There's a certain voice: smooth, agreeable, a little corporate, allergic to a strong opinion. It's not bad writing. It's just nobody's writing. And the moment your readers feel it, they trust it a little less.
That voice is the default, and the default is a choice you didn't make. Tone is the fifth input in the CROFTC framework, and it's the one that decides whether the output sounds like a brand-safe robot or like you.
What Tone actually is
Tone is the personality of the answer: professional, warm, blunt, playful, plain. The model can write in almost any voice, but if you don't pick one, it reverts to the average of everything it has read, which is that polished, agreeable nowhere-voice. Not because it's the best, but because it's the safest bet when it has no idea who's talking.
Tone is also the first thing to drift when you leave it out. You might get lucky on one reply and then watch the next three slide back to corporate, because there was never an instruction holding the voice in place.
Why skipping it costs trust
People connect with a point of view, not a committee. When your emails, posts, and pages all sound like the same faceless assistant, they blend into the wall of sameness everyone is already tuning out. The information might be fine, but the voice is doing quiet damage: it reads as generic, and generic reads as forgettable.
Tone is what makes writing feel human, and it's almost free to add. That's why leaving it out is such a common, expensive mistake.
How to set it well
The fastest way is to name it plainly: warm and direct, confident but not salesy, plain-spoken, a little dry. Even better, tell the model what to avoid, since the default has habits you want gone.
Here's a prompt with no tone:
Write a message announcing we're raising our prices.
And the same request with a tone:
Tone: Honest and respectful, like I'm talking to a customer I value, not managing a crisis. Plain language, no corporate padding, no fake apology. Write a message announcing we're raising our prices.
The first hides behind soft corporate phrasing. The second sounds like a real person being straight with someone they respect, which is the only version that actually holds a price increase together. Same news, completely different reception.
Want your voice built into your prompts? My free bonus pack includes 500+ ready-to-use prompts plus the one-page CROFTC cheat-sheet, so your AI writing sounds like you instead of everyone. Get the free prompt pack.
Tone carries the rest
Tone is the voice the other five inputs speak in. Context, Role, Objective, and Format decide what the answer is and how it's built; Tone decides how it feels to the person on the other end, and Conditions keep it honest while it does. A right answer in the wrong voice still misses. The right voice is what makes people actually listen.
If you change one habit, make it this: never let the model pick the voice for you. Add one line naming how it should sound, and one line naming what to avoid. That's the difference between writing that could be anyone's and writing that's unmistakably yours.
CROFTC is the structure behind every one of the 1,350+ prompts in my series, and Tone is what makes each one sound like you. If you want a running start, grab my free bonus pack: 500+ ready-to-use prompts plus the one-page CROFTC cheat-sheet, so you can stop sounding like a robot and start sounding like yourself.
Tone is the input people fix last and then wish they had fixed first. The other five work the same way, and I put all six in one place, with the prompts already written, in AI Prompt Engineering Bible (7 Books in 1). Seven books in one volume, each input taken apart and rebuilt, with prompts you can copy and adapt to your own work.
The full CROFTC series
This is one input in a six-part framework. Here's the whole series:
Comments