AI Slop Wins the Sprint. Truth Wins the Decade.
AI Slop Wins the Sprint. Truth Wins the Decade.
By Jamie Watters | April 2026
The Collapse That's Already Happening
In 2023, 60% of consumers said they preferred AI creator content. By 2026, that number had collapsed to 26%. That is not a correction. That is a rout.
Kate O'Neill, researcher and author, named this phenomenon the "authenticity premium." It is not a vague cultural feeling — it is measurable consumer behaviour. Studies published in the Journal of Business Research show that when consumers believe emotional content was written by AI rather than a human, they experience something researchers describe as moral disgust. They judge the content as less authentic, show weaker engagement, and are less likely to buy — even when the content itself is objectively identical.
The Nuremberg Institute found that simply labelling content as AI-generated lowers purchase willingness. Not bad AI. Not wrong AI. Just labelled AI.
McDonald's learned this the expensive way. Their Netherlands team produced a technically competent AI-generated Christmas commercial. Audiences responded with comments like "ruined my Christmas spirit." McDonald's pulled the ad. They called it "an important learning." What the audiences felt was the absence of something the AI could not manufacture: a human being who actually cared.
As Becky Owen, CMO of Billion Dollar Boy, told Digiday: "Consumer sentiment is roaring against AI. They hate it. We're in this massive reset."
Why Slop Wins the Sprint
If consumers are rejecting AI content, why is so much of it still working — at least for now?
Three reasons.
Volume arbitrage. Producing 100 mediocre AI posts costs less than producing 5 good human ones. If even 2% convert, the math works. Short-term, quantity beats quality.
Algorithm lag. Search engines and recommendation systems are still catching up. The signals they use to rank content are being gamed faster than they can be updated. The window is open. For now.
Audience inexperience. Most people cannot reliably identify AI-generated content yet. The tell-tale signs — generic phrasing, hollow authority, answers that feel like summaries rather than experiences — are not obvious to everyone. Yet.
Each of these advantages is temporary. The algorithm lag is closing. Consumer AI literacy is growing. And as the volume of slop increases, the signal-to-noise problem compounds until audiences simply disengage entirely.
A University of Florida study published in the Journal of Marketing Research found that AI slop hurts both consumers and professional creators by congesting recommendation systems — making it harder for anyone to encounter genuinely useful content. The people generating slop today are not just degrading their own credibility. They are degrading the ecosystem everyone operates in.
Why Truth Wins the Decade
The authenticity premium is not just a cultural mood. It is structural.
Google's E-E-A-T framework — Experience, Expertise, Authoritativeness, Trustworthiness — is the operational definition of what search engines reward. These signals favour first-hand experience, named authors with verifiable track records, and original data. An AI can summarise existing knowledge. It cannot build a product, encounter a real problem, solve it, and report back with the battle-tested lessons.
AI discovery systems are evolving in the same direction. As AI becomes the primary interface through which people find information, the signals it uses to decide what to surface are becoming the new gatekeepers. And those signals increasingly favour proof over claims.
There is also a structural reason this compounds over time. A landmark study in Nature demonstrated that AI models trained recursively on AI-generated data undergo "model collapse" — outputs progressively lose fidelity and eventually become nonsensical. The implication is stark: AI systems fundamentally need original human data to survive and improve. The people generating slop today are undermining the training data that future AI systems depend on.
This is why I am betting on truth. Not because it is noble, though I think it is. Because the economics point in one direction over any meaningful time horizon.
What This Means in Practice
I am building a portfolio of 20+ products, sharing every metric openly — including the ones that make me look bad. $0 MRR. 0 users. Projects that fail. Decisions that cost me time and money.
That radical transparency is content AI cannot replicate. It requires skin in the game — the one thing no language model possesses.
"When 'perfect' content is free, 'authentic' content becomes priceless."
The sprint goes to the slop machine. The decade goes to the people who were honest about what they were building, what they got wrong, and what they actually learned.
I know which game I am playing.
This is part of the Fool's Gold series. Read the pillar article: The Road to Hell is Paved with AI-Generated Fool's Gold
Building in public at jamiewatters.work