Every naming phenomenon this large produces a joke layer, and AI slop has produced two distinct ones. One is the material itself. The other is the vocabulary people invented to argue about it, which turned out to be more durable.
The content that became the joke
Shrimp Jesus. The founding image of the genre. A photorealistic Jesus rendered in shrimp, shared on Facebook alongside similar material, including animals posed like people and children in impossible situations.
What makes it a meme rather than a failure is the intent. These images are not attempting to be believable. They are built to stop a thumb, and absurdity is the cheapest way to do it. The joke is the format itself, which is why it kept working after everyone understood it.
Italian brainrot. A wave of generated characters and videos with invented Italianate names that spread across short-form platforms and produced its own recognisable aesthetic. Unlike Shrimp Jesus, this one built a genre with recurring characters and a fan base.
The pattern worth noticing is what happened next. Within months, brands adopted the aesthetic to reach younger audiences, which is where a meme's meaning usually starts to drain. Wikipedia's documentation of AI slop records that adoption among advertisers.
The jokes people make about slop
This is the more interesting half, because it became infrastructure for criticism.
The reaction image. “Your AI slop bores me” began as a picture posted by the Artists Against Generative AI Facebook page in October 2025, built on an older meme format and designed for one job: to be dropped into the comments of a generated post as a reply. It later became the name of a viral website.
The efficiency is the point. A reaction image ends a conversation without making an argument, and it lets a criticism travel without the person posting it having to write anything. Within six months a comment-section retort had become a project with thousands of concurrent users.
Microslop. A nickname for one company's AI-branded features and their output. It compresses an entire critique of a product strategy into two syllables, and like all good insults it is funny enough that people use it whether or not they agree with the whole argument behind it.
Slopper. The personification. A pejorative coined in 2025 for someone who leans on generative tools to produce work they would otherwise have done themselves. As a word it is doing something different from the others: it targets a person, which is why it shows up in fights rather than in analysis.
The em dash argument
The most meta of the memes, and the one most often mistaken for a test.
Somewhere in the last few years, the em dash became a widely circulated indicator of machine-written prose. The observation is quoted in academic writing on slop as an example of how recognizable machine style became, with the writer Wesley Yang's remark about it cited in the paper on measuring slop in text.
The problem is that it does not work as a test. Plenty of professional writers use em dashes constantly. Plenty of generated text avoids them. What the em dash meme actually demonstrates is the pressure people feel to make a judgment they have no reliable way to make, and the resulting reach for any available signal.
That makes it a useful thing to have in this list, because it shows the cultural layer doing something the technical layer cannot. There is no working detector, so the culture invented folk tests, and the folk tests are wrong in both directions.
Why the vocabulary outlasted the gags
Look at what survived from the last three years: slop, workslop, slopper, brain rot, Microslop, AI fatigue. Every one of them names a complaint that people had already been making without a word for it.
That is usually what a successful coinage does. It does not introduce an idea. It hands people a handle for an experience they were already having and could not describe efficiently. The reason slop beat AI garbage and AI pollution is the same reason the jokes beat the think pieces: shorter, funnier, and easier to use in an argument.
The joke this site is inside of
There is an obvious trap here, and it would be dishonest to write a culture page about slop without naming it.
A website that explains AI slop, lists AI slop examples, and glosses AI slop vocabulary is structurally identical to the thing it criticizes. The format is the same. The risk is real.
What separates the two is not the topic and not the tone. It is whether the pages contain anything checkable: named sources, specific claims, and a willingness to say when something is contested or unknown. That is the whole argument of this site, and every page is either evidence for it or against it. A glossary entry with a citation does something a generated listicle cannot. A sentence with a number and a source attached can be checked by a reader in thirty seconds, which is exactly what all the slop in the examples above has trained people to stop doing.