The fear going around is that AI will flatten every brand into the same gray voice. Then only the people who already had a name, a team, or a family business behind them will look trustworthy. The first half is right. The second half does not follow.
Sameness is real. Open ten competitor feeds in any trade and count how many sentences could swap accounts without anyone noticing. Most of them. But sameness makes trust harder to fake, which is a different thing from making trust harder to earn.
The paste comes from the prompt
A model asked for a post about your services returns the average of every post about services it has seen. Average has no address. It has no price, no street, no job from last Thursday that went sideways. Anyone can produce it, so nobody can be believed for producing it.
That is a property of the input. When the only thing you hand the tool is a topic, you get the mean of the internet back. When you hand it a transcript of what you told a customer on the phone this morning, you get something with edges.
So the tool is not the problem. Owners who skip the raw material and take the first draft are the problem, and they were doing the same thing with stock photos and template newsletters ten years ago.
Polish used to be a receipt
For a long time, a clean brochure or a well-built site told a buyer that somebody had paid for it. That was weak evidence of anything, but it was evidence. It cost money to look finished, and a buyer could read the cost.
Now looking finished is free. The receipt is worthless, and buyers know it. They are already shifting toward what is still expensive to fake: a stated price, a named type of job, a position you took in March that you still hold in October.
Money does not buy those. A well-funded competitor can hire a team and post daily. It cannot make the posts remember anything unless a person on that team holds the memory and has permission to say true things in public. Most funded brands have neither. Their legal review sands every claim down before it ships.
This is where I think the nepo-baby worry goes wrong. Wealth buys reach, and reach was never the same as trust. A stranger deciding whether to call you does not ask who your parents are. He asks whether you sound like someone who has done this job and will say so plainly.
What you already have
You have sales calls. You have a support inbox that never empties. You have a short list of things you explain to every single customer, in nearly the same words, because the same misunderstanding shows up every time.
That list is the asset. Take a roofer who tells every homeowner the same three things about flashing, why the cheap quote skips it, and what a leak looks like eighteen months later. No model produces that unprompted, and no competitor’s marketing hire can reproduce it without standing on your roof. I made the broader case about a plumber who taught instead of advertised, and the mechanism is the same: you give away the theory and sell the hands. It’s here: https://guruvox.substack.com/p/the-plumber-who-taught-instead-of
The point for this argument is narrow. Your best material is cheap for you and impossible for anyone else to counterfeit. AI sameness raises its value. When every other feed reads like the average, a post with a real number in it stops the scroll by contrast alone.
The advice was priced for a team
“Post every day.” “Be on five platforms.” “Stay consistent.” This advice assumes someone’s full-time job is to follow it. If you run the company and answer the inbox, it reads like a bill you cannot pay.
So you pay it badly. You post for two weeks, a big project lands, the feed goes dark for six, and then you come back with something loud. A stranger looking at your account during those six weeks sees a brand that may have closed. That is often exactly when he is deciding.
The fix is smaller than the advice. One voice, written down. Three offers, explained the same way each time. A rhythm you can hold on a bad week, even if that rhythm is two posts and a few replies. Continuity does more work than volume, because continuity is the thing a buyer can check. He scrolls back and finds that what you said in spring matches what you say now.
Funded brands struggle with this too. Three people answer the same question in three registers by Thursday, and the confident LinkedIn voice turns to hedging in the support thread. A small operator with one head and one voice has a structural advantage there, if the voice survives the busy weeks.
A test for the first draft
Whatever tool or person produces your first draft, run one check before it publishes. Put a competitor’s logo on it. If it still reads as true, kill it. It was never about you.
Then add what only you can add. A price range. A job type you refuse. A mistake customers make before they call. If you have never published your real number, I went into why that filter helps you more than it costs you here: https://guruvox.substack.com/p/the-question-everybody-refuses-to
Some of those additions will make you look blunt to your own industry. Leave them in. The sentence a cautious editor wants to soften is usually the one the buyer believes.
What a desk can and cannot claim
I run a writing desk, so I should say where it stops. We cannot invent your experience. We do not promise reach, rankings, or follower counts, because we cannot see those in the work itself. We will not name you publicly without a written release, and Shabbat and Jewish holidays stay offline unless you say otherwise.
What we can do is hold the memory. Kickoff is a working session on your voice, red lines, channels, and calendar. After that the desk keeps your three offers explained the same way, answers the hard questions without flinching, and posts on the ordinary Tuesdays when you have no time to.
The beige paste will keep arriving in volume. The brands that look real through it will be the ones that remember what they said last week.
If you want to see whether the desk fits your business, book a strategy session with a co-founder.