Talk:AI slop

Did you know nomination

The following is an archived discussion of the DYK nomination of the article below. Please do not modify this page. Subsequent comments should be made on the appropriate discussion page (such as this nomination's talk page, the article's talk page or Wikipedia talk:Did you know), unless there is consensus to re-open the discussion at this page. No further edits should be made to this page.

The result was: rejected by reviewer, closed by AirshipJungleman29 talk 11:25, 28 November 2024 (UTC)

(

• Comment or view
• Article history

)

“Shrimp Jesus” is a commonly cited example of AI slop

• ... that slop emerged in 2024 over “pollution”, “garbage” and “dross” as the preferred term to describe low-quality AI-generated material?

• Source: https://www.nytimes.com/2024/07/24/opinion/ai-annoying-future.html

• • Reviewed: N/a
  • Comment: First own DYK nomination. I like this one because a lot of people contributed (images, categorizations). Feels very fresh.

Created by Jenny8lee (talk). Number of QPQs required: 0. Nominator has fewer than 5 past nominations.

Jenny8lee (talk) 20:39, 4 October 2024 (UTC).

General: Article is new enough and long enough
New enough: Long enough:

Policy: Article is sourced, neutral, and free of copyright problems
Adequate sourcing: Neutral: Free of copyright violations, plagiarism, and close paraphrasing:

Hook eligibility:

• Cited:
• Interesting:

Image eligibility:

• Freely licensed:
• Used in article:
• Clear at 100px:

QPQ: None required.

Overall: Welcome to DYK Jenny8lee, I hope you have a wonderful time here, and I hope I can help facilitate your entry the best I can! Created and nominated within a week, long enough, sourced and neutral. QPQ not required because of the nominator being below 5 nominations. I have some questions regarding the hook, since the original article's wording was as such:
"One increasingly intuitive answer is “garbage.” The neuroscientist Erik Hoel has called it “A.I. pollution,” and the physicist Anthony Aguirre “something like noise” and “A.I.-generated dross.”"
It seems like these terms were used by one or two people, rather than being in use beyond those individuals quoted within the article. I believe it might be a good idea to re-word the hook(something along the lines of "AI slop has been referred to as "garbage", "pollution", and "dross"), or find something else altogether. The quotes within the lead of the article should be attributed to their sources as well. In addition, I'm not entirely sure whether the AI generated image is free (regarding the copyright of a derivative work, this is a fairly new policy as well). I'm going to ping @Theleekycauldron: to see what they think regarding the matter. Ornithoptera (talk) 02:34, 12 October 2024 (UTC)

• @Ornithoptera: I think I'd go further than you to say that the Times source doesn't even verify that "slop" is the preeminent term! Since all AI imagery is considered public domain by Commons, public domain derivatives are a-okay. theleekycauldron (talk • she/her) 18:15, 14 October 2024 (UTC)
• @Ornithoptera: Thanks for the feedback. How do I change the hook once it has been submitted? Jenny8lee (talk) 19:24, 20 October 2024 (UTC)

• @Jenny8lee: To propose a new hook, start a new line with ALT1 (in bold) below, then put the text of the hook next to it. Z1720 (talk) 23:15, 29 October 2024 (UTC)

• ALT1 that thousands of people showed up in Dublin, Ireland for a non-existent Halloween parade due to an article on an AI-produced website in what became a viral example of AI slop in the physical world.

Source 1: https://www.independent.co.uk/news/world/europe/dublin-fake-halloween-parade-ireland-ai-advert-b2639505.html Source 2: https://www.nytimes.com/2024/07/24/opinion/ai-annoying-future.html — Preceding unsigned comment added by Jenny8lee (talk • contribs)

Per WP:DYKTRIM, I suggest ALT1a: ... that a viral example of slop prompted thousands of people to visit Dublin for a non-existent Halloween parade?--Launchballer 13:50, 7 November 2024 (UTC)

@Jenny8lee and Theleekycauldron: Went to tick off ALT1a (since it's a derivative of Jenny's ALT1 with no additional information); however, I noticed that only one of the references actually uses the word 'slop'. And having read the article, the vast majority of this shouldn't be in this article. Most of this deserves to be in the artificial intelligence art article.--Launchballer 14:24, 17 November 2024 (UTC)

@Jenny8lee: Please address the above. Z1720 (talk) 13:49, 24 November 2024 (UTC)

'Unwanted' in the Definition is Too Subjective and Should Be Reconsidered

The second paragraph contains this sentence quoted from a New York Times article: "shoddy or unwanted AI content in social media, art, books and, increasingly, in search results."

The word "unwanted" can be subjective, as different people may have varying reasons for rejecting AI-generated content. For instance, someone who dislikes AI on principle might consider all AI-generated content unwanted, even if it is high quality and useful to them. Likewise, unwanted does not necessarily mean low quality—some AI-generated content may be undesirable for reasons unrelated to its accuracy or overall merit.

I recommend paraphrasing the quote to better reflect an objective characterization of AI-generated content, such as "low-quality or disruptive AI-generated content in social media, art, books, and search results." This maintains the intended meaning while avoiding subjective terminology. 129.222.112.4 (talk) 07:13, 24 January 2025 (UTC)

Yeah but even if its high quality ai its still unwanted. 2600:6C40:7400:3D7:63E:F004:3708:46A7 (talk) 16:12, 10 June 2025 (UTC)

AI slop, n. Content that the user of the term assumes to be AI-generated, regardless of whether that is actually the case.

Study after study after study after study after study after study appears that demonstrates that humans are actually shit at distinguishing AI from human content, yet people insist that "they can always tell" – a classic example of Dunning-Kruger-like unwarranted confidence. Turns out that humans may think they are intelligent, but they're not actually that intelligent, just hallucinating their correctness, because they are sceptical towards everything but themselves. --Florian Blaschke (talk) 20:58, 15 December 2025 (UTC)

Isn't "slop" definitionally unwanted? The sources also pretty much all use the term derogatorily or are quoting people using it negatively. Catboy69 (talk) 21:05, 15 December 2025 (UTC)

If the AI content was wanted by someone, then that person wouldn't call it "AI slop". Brynn Who Likes Editing (talk) 09:03, 12 July 2026 (UTC)

you forget to mention a ad in Brazil against CP (IIRR)

a group of anti-CP or goverment?[sic] in brazil made a ad to combat child M/R/SA and improve. (if you know what that means) however, they used AI Don't go offend those dudes however i remember their number "127" which was a telephone number 2804:3690:8001:57BE:68D2:613D:14BC:CB58 (talk) 16:59, 26 May 2025 (UTC)

I want delete myself before you ignore it 2804:3690:8001:57BE:68D2:613D:14BC:CB58 (talk) 17:00, 26 May 2025 (UTC)

what consarn _{(grave) (obituary)} 17:46, 26 May 2025 (UTC)

I think he meant to say he wanted to off himself lmao ~2026-41156-35 (talk) 15:03, 22 July 2026 (UTC)

Origin of the term

The use of "slop" as a suffix referring to low-quality media originates from the term goyslop.

https://news.ycombinator.com/item?id=40648241 95.205.117.189 (talk) 14:14, 28 May 2025 (UTC)

The article reframing this as a conscious choice among journalists is so disingenous. I suppose they were trying to sanitize the term from its blatantly antisemitic origins, which is quite embarassing for the leftist anti-AI crusaders. ~2026-30125-14 (talk) 20:49, 4 August 2026 (UTC)

Nope. It originated from a comparison to fast food, for which the term has been found in use in 1993. In solidarity, PokemonPerson (talk) 01:34, 5 August 2026 (UTC)

Request to add category "In comedy"

i would like to add a comedy section that goes like this

Overly broad scope?

Like, obviously the reaction to most (pretty much all obvious cases with the least bit of general visibility) AI stuff will be negative, but I feel we do need to have some editorial discretion between the content that does fall under the "definition" of slop vs just anything that gets described as such. It's a buzzword that gets thrown around, and while the phenomenon of enshittification via AI needs to get documented, the page currently several sections which are just unrelated examples whose only connection is that someone called them "slop". ~2026-32512-46 (talk) 06:34, 6 June 2026 (UTC)

Ironically the enshittification Wikipedia article has the same issue; it's a term with a specific technical meaning about the ecosystem of individual and business users of an online service, but which some people (and sources) use more broadly to mean "became more shit".

It may be that the definition of slop is changing, in which case we should probably split out a section on the specific use case of low quality AI content "produced in high volume" to overwhelm or exploit a platform. Belbury (talk) 07:40, 6 June 2026 (UTC)

It sure seems like in recent years there's a fast pipeline online for new English words to appear that initially mean something very specific, but then rapidly transform into a general label for "thing that I don't like." Slop, scam, grift, shill, and so forth. It's a bit of a dilemma for something like Wikipedia, where the title itself has a rapidly shifting meaning that can invalidate the article out from under it. Bryan Derksen (talk) 23:14, 5 July 2026 (UTC)

"AI brainrot" listed at Redirects for discussion

The redirect AI brainrot has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2026 July 10 § AI brainrot until a consensus is reached. Jay 04:10, 10 July 2026 (UTC)

Definition is too narrow: remove restriction to attention economy

The definition in the first setence starts off well:

> is digital content made with generative artificial intelligence that is perceived as lacking in effort, quality, or meaning,

but then incorrectly ANDs this together with a second condition:

> and produced in high volume as clickbait to gain advantage in the attention economy, or earn money.

This second condition only describes some slop, not all; the first part of the definition sufficed without the restriction. Indeed the next sentence goes on to make the point that the second part of the definition was making:

> It is a form of synthetic media usually linked to the monetization in the creator economy of social media and online advertising.

so the second part of the definition can probably simply be deleted (the page seems to be locked though so I don't think I can). Myrmornis (talk) 20:40, 1 August 2026 (UTC)

Noting that User:Zowayix001 has since softened this "AND" definition by adding the word "usually" to it, as part of a body edit about it being slop when a politician leaves an AI prompt in their speech. Belbury (talk) 13:30, 7 September 2026 (UTC)

Cambridge Analytica

The Cambridge Analytica site has been resurrected (under new ownership) and seems to consist primarily of AI slop. Thoughts on whether this warrants inclusion in this article? Here's more info on it: https://futurism.com/artificial-intelligence/cambridge-analytica-ai-slop-farm

It's an interesting story, and also bridges the gap into things like SEO content and link farms, which are not currently covered here. Dflovett (talk) 01:09, 3 August 2026 (UTC)

"AI circlejerk" listed at Redirects for discussion

The redirect AI circlejerk has been listed at redirects for discussion to determine whether its use and function meets the redirect guidelines. Readers of this page are welcome to comment on this redirect at Wikipedia:Redirects for discussion/Log/2026 August 21 § AI circlejerk until a consensus is reached. /over.throws/^✎ 17:35, 21 August 2026 (UTC)

Regarding ai meme removal

@Belbury, I noticed that you removed the post meme/hoax. To help build some consensus I am taking this matter here. While I think it's inclusion is fine, as haveing an AI meme poster was the introduction that many people had to Ai slop. Thoughts? Issac I Navarro (talk) 21:49, 21 August 2026 (UTC)

The content doesn't seem to match the definitions we're using of slop, or to use the term in sources. An AI generated Hitler poster that was apparently a one-off and sufficiently accurate to cause "outrage from users who believed the image to be a genuine Disney production" sounds more deepfake than slop. Disney asking Bing AI to stop people from making Disney posters of their own pets sounds like a case of many individual people creating one image of reasonable quality in good faith, rather than people cynically making knowingly low quality content in bulk.

Broadening the range of this article to include all news stories where AI-generated images are criticised in some way seems like a classic WP:COATRACK, and something to be careful about. Belbury (talk) 10:57, 22 August 2026 (UTC)

Yes, this Hitler/Holocaust image is an absolutely unnecessary image. It fails pretty much every standard for choosing images for encyclopedia articles. It does not match the topic, it is not clearly relevant to the topic, it is not significant to the topic, and it is a potentially offensive image that adds no necessary information about the topic. This should not have been added once, let alone restored after removal. I will be the next person to remove it. Asparagusstar (talk) 18:15, 22 August 2026 (UTC)

Proposal for a different lead image

I would like to propose replacing the current lead image of this article. There are many examples of AI slop and I wonder whether using a religious figure, which some people may find offensive, is the best example. Of course on articles about religious topics, it is necessary to show religious images, but in this case there are other examples that could be used. The depiction of a religious figure combined with a grotesque animal is perhaps a bit insensitive.

I would suggest these alternatives as the lead image:

• File:AI-generated veteran birthday sign image.png, which already appears further down in the article.
• File:AI generated woman with extra legs.jpg, which shows a person with too many limbs, a good example of AI slop.

I'm of course open to other ideas as well. Asenoner (talk) 03:14, 3 September 2026 (UTC)

The three "Shrimp Jesus" images are from reliable sources (like The Guardian, Business Insider) who label them as AI slop. The two images suggested look like images made by wikipedia editors. We should prioritize reliably sourced images in the article, and definitely for the lead images. So, yes, open to replacing the current lead images with alternatives from reliable sources. The user images should be removed the article, as they are likely not actually AI slop. Asparagusstar (talk) 12:58, 3 September 2026 (UTC)

as a heads-up, "religious insensitivity" isn't a thing to worry about here, with wikipedia not being censored and all that (which is why showing images of muhammad is fair game). really, the people who actually buy into this slop when it comes to religion are more insensitive anyway, worshipping false idols made with the mangled remains of fanart of their sky daddy of choice, but that's besides the point

in this case, the merman lottery loser jesus images, as ones supported and explicitly described as ai slop by sources, are then fine and dandy for presenting the general nature of ai slop consarn _{(talk beige) (count ribs)} 13:18, 3 September 2026 (UTC)

i should note that this doesn't mean that i support or oppose their inclusion as lead images, just that "it could bum christians out" isn't a reason to oppose

also shrimp are cute, even if nature hates them consarn _{(talk beige) (count ribs)} 13:24, 3 September 2026 (UTC)

Textual markers and the em dash evidence

Hi all. The article is strong on images, platforms and the economics of slop, and the Kommers et al. properties are well summarised. What's missing is the textual side: which features actually mark machine prose, and the fact that the best-known marker has now been measured and doesn't hold.

Disclosure: I edit at Silent Room, a magazine about AI and writing, so per WP:COI I'm raising it on talk.

Suggested as a short subsection:

Discussion of textual slop has produced an informal catalogue of markers: antithetical framing ("not X, but Y"), triads where two or four items would serve, hedging openers such as "it's worth noting that", a recurring lexicon, and closings that are rhetorically weighted but content-free with the caveat that none is an error in itself, each appearing in competent human writing, and that what distinguishes the machine version is frequency rather than the device. Such markers are also unstable, because models are tuned as each becomes recognisable: in "How to spot AI writing" (July 2026), The Economist compared 55,940 sentences of its own journalism with versions generated by four models on the same prompts and found ChatGPT now uses the em dash less often than human writers, retiring the most widely cited tell. Reviewing the catalogue against that finding, the magazine Silent Room argued that detection guides are structurally behind the thing they describe, since a marker disappears from model output roughly as soon as readers learn to look for it. Related work has examined the review burden this creates: "AI Slop and the Software Commons" (2026) frames the asymmetry as content that is cheap to produce and expensive to assess.

The Economist finding seems worth the article on its own, the catalogue is the part I'd expect people to want trimmed. Enthusiast1874 (talk) 11:07, 8 September 2026 (UTC)

In event listings - AI flyers

recently there has been an usage of AI flyers around small businesses and local events that already had human made flyers, has anyone been able to find new sources mentioning this besides social media? Liamleano (talk) 16:34, 21 September 2026 (UTC)

https://www.404media.co/we-are-living-in-a-chatgpt-flyer-pandemic/ is a secondary source covering the social media reactions, which could be quoted.

There have been a few articles about individual venues banning AI-generated band posters:

• https://www.sfchronicle.com/entertainment/article/thee-stork-club-ai-flyer-ban-21044992.php
• https://www.bendsource.com/music/resisting-the-ai-assist/
• https://www.cbc.ca/news/canada/nova-scotia/canada-s-music-scene-sees-pushback-against-ai-generated-artwork-9.7324581

Belbury (talk) 16:48, 21 September 2026 (UTC)

rat image

so do we really need to have an image of a rats ballsack to show what "ai slop" is? i think we got the point LoganStylus (talk) 22:44, 30 September 2026 (UTC)

No, but it is useful for illustrating AI slop in scientific papers. We could use another image though, especially regarding Wikipedia:GRATUITOUS, you're welcome to find and add one. In solidarity, 💫ΩmegaMantis💫^(he/him) ❦blather | ☞spy on me 01:14, 1 October 2026 (UTC)