76% of small businesses use AI. Of small businesses using AI, 93% say it has had a positive effect on the business.[1]

But 65.5% of business owners say they worry AI will make their business feel less personal to customers.[2]

That's a lot of people who think AI makes us look like we are cutting corners. It doesn't have to but it often does.

There's a phrase going round at the moment. ai;dr

It comes from tl;dr, which is old internet shorthand for "too long; didn't read." You skip the long thing and go to the summary.

ai;dr means "AI, didn't read." Someone decides a post was written by a machine, so they don't want to read it.[3] It's funny. It is also expensive if it happens to you.

Last month, on 30 July, LinkedIn added a button to every post so people can flag it as "seems like AI slop." The company had already been blocking billions of automated comments.[4]

So people are annoyed by AI content. How is that sentiment shifting? How does it impact your revenue?

Reading the Jerk Layers this week

Position

77% of US small businesses use AI regularly.[8] Only 17.7% have ever paid for an AI subscription.[5]

JPMorgan counts direct subscriptions. It does not count AI built into Canva, your CRM, or your email tool, and the authors say the figure may understate use.  Even so, most people writing with AI are writing with a free model which is the weaker tier.

33% of consumers say AI worsens their perception of a brand. 16% say it improves it.[6]

60% say seeing the word "AI" in a company's messaging puts them off.[7]

Velocity

That 77% was 48% in July 2024.[8]
29 points in eighteen months.

Meanwhile consumer preference for AI creator content over traditional creator content fell from 60% to 26% over three years.[9]

More businesses using it. Fewer customers wanting to see it. Those two lines are diverging.

Acceleration

Bank records show this better than surveys do. JPMorgan Chase tracked what small businesses actually paid for, 2019 through 2025.

1.7% had paid for an AI tool in 2019. 4.5% by 2022. 8.3% by the end of 2023. 11.6% in 2024. 17.7% by December 2025.[5]

The yearly increase went 0.9 points, then 3.8, then 3.3, then 6.1.

Last year was the steepest yet. Paid adoption is accelerating. But it is not a majority yet.

On the customer side there is no series long enough to measure acceleration. Only direction, and the direction is down.

Jerk

Then came the button on LinkedIn.

This is a new feature, not a fad. Platforms don't ship new features unless they are likely to drive revenue or cut costs. It likely means they were spending too much on the consequences of AI slop.

There is no series long enough to compute the third derivative on the customer side. A platform building infrastructure is the early qualitative signal that AI resistance crossed a threshold.

52% of consumers say they would stop buying from a brand after an inauthentic experience. 91% expect brands to disclose when they use AI.[10]

And the penalty is not only for bad work. Researchers played people music. The listeners liked it. Then they were told it was AI-generated, and their willingness to pay went down.[11]

So AI slop pushback has moved out of opinion and into how people choose.

Beyond quality  

If your content looks like everyone else's you are indistinguishable to AEO. (AEO is answer engine optimization. How you get quoted when someone asks an LLM, or when Google writes the summary.) An engine choosing between four identical descriptions returns one of them. There is no reason for it to be yours or to cite your page.

AI answers tend to favor original content that is easy for them to read.

If your content is slow to load it is less likely to be pulled.

If it reads like everyone else's, it may end up in the summary without your name and without a click.

2 admissions

I use Claude to help research and craft my emails.

And if the word AI puts off 60% of customers, then some of what I've named my own work has exactly the problem I've just described. I'll be rephrasing the descriptions on what I do.

You might want to look at your own content this week.

Your five-minute practice this week

Prompt your favorite LLM.

1. The check.

Here is a page from my website. Return only the sentences that fail either test. For each one, tell me which test it failed, and quote the exact words that triggered it. Test one: does it read as AI-written. Test two: could this sentence appear, word for word, in a competitor's version of the same piece. Don't rewrite anything. Don't comment on the sentences that pass.

Reading what comes back.

• Failed test one only. Your content is unique but the writing is not. A writing problem, and the steps below fix it.

• Failed test two only. You wrote it yourself and it still could have come from anyone in your category. That's a question about the business rather than the post.

• Failed both. Start with the writing. Then take the second list seriously.

2. Don't let it rewrite.

Not yet. Instead ask:

For each line you flagged, ask me one question only I could answer. A number, a name, a date, something that happened, a unique approach to the problem, a different point of view. One question per line, ten questions maximum. Don't write anything yet.

3. Answer roughly.

Fragments are fine. Don't tidy them up.

4. Rebuild from your answers.

Now rewrite only the flagged lines, using my answers. Use my words where I gave you words. No new adjectives. No new transitions. Leave every unflagged line exactly as it was.

5. Repeat step 1. Optional. One pass is enough to see the problem.

I suggest you do this on the page that matters most. Usually it’s the one you send people to, not the homepage.

What makes the writing on your page yours is both style and the things you know that others don't.

Your unique value proposition is a big part of what drives your visibility in AEO.

Ask an AI to recommend a business like yours. Then ask again tomorrow. Different answer. Unless you’re entirely unique.

Anyone who tells you they can put you in that answer is wrong. Nobody can. You can only change the odds. Or buy an ad next to the answer.

This is why I don’t offer to maintain websites. Answer Engine Optimization (AEO) is changing too fast for guarantees. What I can help you with is increasing the odds of AEO discovery.

That's what the Increase Your AEO Odds Workshop is for. Tuesday 1 September, 1–3pm ET, live on Zoom, $297.  

In two hours we will cover:

• The questions your buyers actually type, in their words, not yours.

• The one thing you do that your competitors cannot claim, written in a sentence you can paste.

• Your own pages checked against what answer engines can read.

• A short list of what to fix first including the types of consistency and structural changes you should be doing.

• Where the answer engines are actually pulling from right now, and what changed this quarter.

You do the work in the room. You leave with content ready to publish.

Best wishes,

Rose

Sources

1. Goldman Sachs 10,000 Small Businesses Voices, surveying 1,256 small business owners across all 50 states, DC and Puerto Rico, fielded 27 January – 4 February 2026 by Babson College and David Binder Research: "Survey: Small Businesses Embrace AI — But Need Training and Support to Fully Harness It"

2. UPrinting / Pollfish survey of 1,000 US adults including small business owners, managers and entrepreneurs, reported by Stacker: "1 in 4 business owners say AI is costing them clients: Here's what that means for 2026"

3. Fast Company, "'AI; didn't read': AI;DR is the new TL;DR". See also Futurism, "There's a Grim New Expression: 'AI;DR'"

4. TechCrunch, "LinkedIn adds a button to report AI-generated 'slop'", 30 July 2026. Automated-comment figure from Fortune, "LinkedIn adds a 'seems like AI slop' button after blocking billions of automated comment attempts"

5. Wheat, Christopher, Chi Mac, and Andrea Passalacqua, "Understanding the use of AI among small businesses", JPMorgan Chase Institute, 14 April 2026. Based on Chase Business Banking transaction records for 4.6 million small businesses, 2019–2025. Measures firms that have ever paid for an AI service; the authors note the method excludes free tools and embedded AI features and "may understate actual usage"

6. Clutch, "AI in Branding: Why 33% of Consumers React Negatively", reporting a Clutch survey of 408 consumers, June 2026. Page updated 14 July 2026

7. WordPress VIP, "Future of the Web 2026", surveying 2,000 respondents (800 enterprise decision-makers and CMOs, 1,200 US adults), fielded April 2026. Reported by TechCrunch: "Sixty percent of U.S. consumers say 'AI' in brand messaging is a turnoff, survey finds", 16 June 2026

8. Intuit QuickBooks, "2026 AI Impact Report", surveying more than 34,000 US small businesses in partnership with the University of Chicago, alongside anonymised data from more than 5.3 million QuickBooks businesses across the US, Canada, UK and Australia. US readings of 48% (July 2024) and 77% (January 2026)

9. Billion Dollar Boy, surveying 6,000 consumers, creators and marketers across the US and UK, reported by Communicate Online: "97% of marketers use AI, yet most consumers still prefer human-made ads"

10. Emplifi, reported by EMARKETER: "Shoppers aren't impressed by AI-generated marketing", 1 May 2026.  

11. ProMarket, "Consumers Prefer AI Music Until They're Told It's AI", 4 May 2026