Growth Marketing

When content converges to the mean: volume is not a strategy

On AI slop, what platforms are already doing about it, and what actually separates content that works from content that doesn't.

Lola Rodríguez Lola Rodríguez · · 7 min read
TL;DR

"Slop" was Merriam-Webster's 2025 word of the year, and platforms are already filtering it out. When everyone uses the same AI tools, output converges toward the mean instead of standing out — the differentiator is editorial judgment, not volume.

In December 2025, Merriam-Webster chose "slop" as its word of the year. The official definition: digital content of low quality produced in large quantities by means of artificial intelligence. Not a new word. In the 18th century it meant soft mud. In the 19th, food waste, what you feed pigs. Today it describes what's flooding everyone's feeds.

What Merriam-Webster didn't say explicitly but left implicit in the choice: the word won because people are using it. And they use it because they recognize the phenomenon. There's an active cultural rejection of content that feels generated without judgment, without a real perspective, without anything worth saving.

That's a market signal. And for anyone who produces content professionally, it's worth reading carefully.

How we got here

The pattern is familiar. When a tool drops the marginal cost of producing something to nearly zero, the logical response is to produce more. It's always been this way with technology. The printing press meant more books. Digital recording meant more music. Social media meant more opinions. And each time, the initial volume surge was followed by saturation that forced a rethink of what actually mattered.

With AI the cycle accelerated. What took decades with other technologies took months here. Large language models reached marketing and content teams in 2023, and by late 2024 there was already enough low-judgment content out there for the phenomenon to have its own name, Wall Street Journal coverage, and CNET articles declaring that AI slop had turned social media into an antisocial wasteland.

The problem isn't technical. The texts are grammatically correct. The images are clean. The videos have good resolution. The problem is they're indistinguishable from each other. And in an environment where attention is the scarce resource, indistinguishable is the same as invisible.

The platforms responded

When a phenomenon reaches the scale of slop, platforms have no choice but to act. Not out of ethics, but survival: their business depends on users wanting to stay.

In October 2025, Pinterest launched specific controls letting users filter AI-generated content from their feeds. The platform had received months of complaints. Interior design, fashion, and art communities, which are Pinterest's core, were being emptied of human content. Users blocked accounts without the feed changing because slop was coming from everywhere. The pressure was enough for Pinterest to introduce the anti-AI toggle, available in categories like beauty, art, fashion, and home decor.

In January 2026, YouTube CEO Neal Mohan published his annual letter to the creator community and put slop at the center. "The rise of AI has raised concerns about low-quality content, aka AI slop," he wrote. YouTube announced it would strengthen its systems to fight spam, clickbait, and repetitive uploads, while simultaneously expanding AI tools available to creators. The irony is deliberate: the goal isn't less AI, it's AI with judgment.

What these responses reveal isn't that slop has won. It's that platforms are building systems to detect it and reduce its distribution. YouTube's algorithm isn't neutral. Pinterest's feed isn't neutral. They're being actively calibrated to penalize content that lacks signals of authenticity, original perspective, or real value to the reader.

That changes the calculus for anyone producing content.

The problem with optimizing for volume

For years, the dominant advice in content marketing was consistency. Publish often, across more channels, with more frequency. Volume was the strategy because production cost was the real constraint. Teams that could publish more had an advantage over those that couldn't.

That constraint is gone. And with it, the advantage of those who could only produce more.

When everyone has access to the same tool and optimizes for the same objective, output converges. Not toward excellence, but toward the mean. An industry executive described it in Marketing Dive with precision: output is trending toward the median, all the content is merging to look very, very similar.

Convergence toward the mean isn't just an aesthetic problem. It has practical consequences for how distribution works today.

What platforms did about AI slop
PlatformResponseWhen
PinterestToggle to filter AI content from feeds (beauty, art, fashion, home decor)Oct 2025
YouTubeStronger anti-spam / anti-repetitive-upload systemsJan 2026
Google SearchE-E-A-T ranking adjustments favoring real experienceOngoing

The AI models that answer questions, recommend products, and cite sources, which are increasingly the first point of contact between a brand and a potential buyer, tend to cite content with verifiable authority, traceable data, and original perspective. Content that paraphrases what already exists on the internet doesn't accumulate authority. It dilutes it. Slop doesn't just fail to appear in Perplexity or Claude responses. It actively competes for space with content that has quality signals, and loses.

In traditional organic search, the effect isn't neutral either. Google has been adjusting its systems to identify useful content versus content designed to rank without adding real value. The adjustments of the last two years explicitly targeted generic content and elevated content with real experience and original perspective, what Google calls E-E-A-T: Experience, Expertise, Authoritativeness, Trustworthiness.

What actually differentiates now

There's a real tension worth naming: the same tools that enable slop also make it possible to produce quality content faster than ever before. The problem isn't the tool. It's the judgment applied when using it.

The difference between content that reads as slop and content that doesn't is almost always in the last step. The one that requires someone who knows how to distinguish. Who has a real perspective on the topic. Who can evaluate whether what the tool generated is worth publishing, and why.

That's not a technical problem. It's an editorial one. And it's precisely where AI still can't replace someone with real experience in the subject.

A paper published in January 2026 by Cody Kommers and five other researchers identified three properties that characterize AI slop: superficial competence, asymmetric effort, and mass producibility. All three are resolved with the same antidote: a point of view that doesn't come from a generic prompt. Real data that required research. First-hand experience that doesn't exist in any training corpus because it's yours.

The authenticity paradox at scale

There's something that doesn't resolve easily and is worth leaving open.

The argument in this article points toward editorial judgment as the real differentiator. But editorial judgment doesn't scale the same way an automated content pipeline does. One person with a real perspective can produce ten good pieces a month. Not a hundred. Not a thousand.

The teams resolving this well aren't automating judgment. They're automating everything that doesn't require judgment and protecting the time of whoever has it so they can apply it where it matters.

It's not a perfect solution. But it's more honest than saying volume is still the strategy.

/ faq

Frequently asked questions

What is AI slop?

Merriam-Webster's 2025 word of the year: digital content of low quality produced in large quantities by AI. A 2026 research paper defines it by three properties: superficial competence, asymmetric effort, and mass producibility.

Are platforms actively filtering out AI slop?

Yes — Pinterest launched an anti-AI toggle for feeds in October 2025, and YouTube's CEO named slop directly in his January 2026 creator letter, announcing stronger anti-spam systems.

Why doesn't publishing more content work as a strategy anymore?

When everyone has access to the same AI tools optimizing for the same objective, output converges toward the mean instead of standing out. Volume stopped being an advantage once production cost stopped being the constraint.

What actually differentiates content now?

Editorial judgment — a real point of view, traceable first-hand data, and the ability to evaluate whether AI-generated output is actually worth publishing. AI models increasingly cite content with verifiable authority over generic paraphrased content.

Lola Rodríguez
Lola Rodríguez
Growth Manager & AI Applied at Cronuts Digital

I write about growth marketing, AI, and project management — from inside the work, not above it.

Connect on LinkedIn ↗
← back to all posts more about me ↗