Slop as a vector of private interest: abstraction, songwriting, and extraction after the public/private divide

“Slop” typically describes the kitschy, formulaic output of generative AI. It can also be a useful term to describe the form of abstraction tools like ChatGPT and Sunno use to create those tasteless outputs. This type of abstraction creates a generic profile into which a variety of specific stylistic determinations could be slotted, depending on the context.

LLM outputs are the result of probability models that can identify a general “vibe” or profile but which can’t nail down the details. For example, when ChatGPT was first released, I asked it to write a biography of me, the scholar Robin James. It correctly ascertained that I went to a PhD program in continental philosophy, but it could never correctly identify which one: repeated querying had it vascillate between Penn State and Vanderbilt – both strong continental programs and I almost did go to Vandy – but it never hit on my actual alma mater, DePaul. It got the generic vibe right, but from its perspective Penn State, Vandy, and DePaul were all interchangeable examples of that vibe. Slop embodies the same kind of probabilistic modeling that LLMs use: it creates a generic profile or orientation that any particular set of qualitative features or determinations can be plugged into. Any genre of music can be chilled out by toning down its sharpest edges, just as any space can be made to appear liminal by vacating evidence of human presence.

Slop can be automated, but it can also be the product of actual human labor. I have literally written the book about how people in general and pop culture specifically have learned to perceive and think like AI; there I argue that “vibes” are the qualitative, vernacular version of the kind of probabilistic thinking behind technologies like LLMs and recommender algorithms. Calling this sort of thinking “slop” emphasizes its genericness and abstraction, and that’s important to do because slop assumes a different kind of relationship between general and particular than is assumed in most of the political philosophy grounding US law.

Classical liberalism imagines all persons to be formally equal before the law; difference in things like race, class, gender, or religion are all considered private phenomena and they are not admitted into the public sphere/civil society, which is where all experience the law’s equal protection. This is why, for example, US courts can only consider things like gender or racial identity in certain conditions of “strict scrutiny”; otherwise justice is blind. As Hegel explains in the Phenomenology of Spirit, “The universal being thus split up into a mere multiplicity of individuals, this lifeless Spirit is an equality, in which all count the same, i.e. as persons” (290). Every one person is an empty identical token of the same equal, universal type. This moment in the Phenomenology recapitulates its opening gesture, where the dialectic of sense-certainty exhibits a particular sort of tension between universal and particular: each “this” or “now” can represent a highly defined and specific point in space or time because it can equally well represent ANY defined point in space or time. This is the “When is now, now?” Moment in the movie Spaceballs. “Now” can be any NOW because “now” itself has no determination other than being an individual moment in time. Hegel calls this a “negative universal” (290). The same is purportedly true with the liberal subject: in theory every person can be equal to any and all others because they have no distinguishing determinations beyond the fact of their being one individual: “the formalism of legal right is thus by its very nature without a peculiar content of its own” (291). Of course, feminists and critical race theorists have famously pointed out that this empty abstract individuality is in fact normatively white cisheteromasculine, and that white women and people of blue were barred from personhood precisely because their “private” differences were incompatible with civil equity. Classical liberalism’s public/private split allowed this form of abstraction to create the appearance of political equality behind the practice of civil segregation.

Slop uses a different method of abstraction from particular to general. Instead of creating a binary division between particular and general, slop emerges when the kind of probabilistic reasoning AI uses sifts through a vast collection of particular data points in order to find the most likely orientation of the most contextually-relevant of those data. Justin Joque describes this form of probability as a theory of “situated likelihood.” For example, LLMs look for the most likely arrangement of words given the context of the prompt they have been fed. This is not a void of difference, but a profile according to which some particulars are more well-suited than others. In the above example of ChatGPT guessing my PhD alma mater, in the context of all the publicly-available data about my scholarship, Vandy or Penn State are more well-suited to the emerging profile it built of me than say NYU or Michigan (both heavily analytic programs). This style of probability shifts through troves of data about private difference in order to infer the most likely profile of the phenomenon in question, whether it’s where the scholar Robin James got her PhD or how to vibe code an agent to make differ reservations for me. A variety of different sets of particulars can be slotted into this general profile, but not any or every set of particulars: the whole point is that some are more likely, more aligned, than others. 

Context is like a Russian nesting doll, and these practices of profiling are situated in broader contexts that shape things like the association between Blackness and criminality or masculinity and job performance. As Roland Meyer put it, “AI #slopaganda is not so much about fakes or disinformation. It’s about aligning the world with your preformulated descriptions of it, making reality disappear behind its formulaic image.” As I argue in Good Vibes Only, the difference between an actionable AI output and a so-called “hallucination” is that the former is aligned with existing biases, whereas the latter is not. The opposite of enlightenment disinterested universality (which naturalized the class interests of the powerful under the guise of formal equality), slop-straction explicitly naturalizes the class interests of the most powerful as such. Slop is a vector of private interests. These profiles that can be populated by a range of different stylistic options create the appearance of post-demographic diversity by allowing for a range of different likelihoods: jazz can be chill, country can be chill, rock can be chill, pop can be chill, even “agit-slop” videos of ICE detainees can be chill, etc. However, that superficial diversity in outputs neveretheless maintains the broader alignment of the private interests that organize so much of our patriarchal racial capitalist world. This style of abstraction creates a profile that makes some individual preferences more likely and more accessible than others. Slop is a vector of private interest that privileges the interests of those already most aligned with power. 

***

Alex Warren’s “Ordinary” and Ella Langley’s “Choosin Texas” are slop in this sense: it’s not a judgment about their aesthetic quality, or a claim that they are made with AI, but a description of how they approach songwriting. In summer 2025 and 2026, respectively, these songs achieved unprecedented dominance of the Billboard Hot 100. Unlike the long tail of “Mr. Brightside,” which has spent over five years somewhere on the UK top 100, these songs spent months and months at #1, “Texas” breaking every record for the longest reign at #1. Each of these songs exhibits the profile of a song that is most likely to perform exponentially well on the Hot 100 (‘to scale’ if you want to use VC-language; another defining feature of slop is that it scales in value): it’s playable across a variety of radio formats, thus boosting airplay numbers, it is Spotify-core enough to work well as background music on a lean-back listening playlist, and it’s good for soundtracking personal videos on social media. Think of it like a song having a “portfolio career”: no single market will generate enough plays to put the song on the top, but putting a bet in every pot possible will. This profile depicts the private interests of the recording industry, sketching out three data points (airplay, streaming, social) that constitute the vector for chart success today and synergizing them for scale. “Slop” in this sense of the term refers to the fact that these songs have the profile of a scalable vector of IP.

These songs’ aesthetic profile is also aligned with the broader interests of our very imperfect and unequal world: songs like 24KGoldn’s “Mood” (eight weeks at #1 on the Hot 100, first track to hit #1 on that chart, Hot Alternative, Hot Rock & Alternative, Hot Rap on the same week) and Gaga & Bruno Mars’ “Die With A Smile” (Spotify’s most-streamed song of 2025) exhibit similar profiles, but they have not had such exponentially scalable success because…they involve people of color and musical styles associated with them. I’ve written before about how Bruno Mars’ racial identity has prevented him from appearing “post-genre”, another way of framing the flexibility and adaptability needed for sloppy scalability. “Ordinary” and “Choosin Texas” ooze white cisheterosexuaity, aligning these songs with the private interests of the ruling classes. It is this broader alignment – not just with the interests of the recording industry but also the interests of patriarchal racial capitalism – that allows Warren’s and Langley’s songs to scale in ways 24KGoldn and Gaga/Mars cannot. 

***

”Cool” was a response to classical liberalism’s construction of the person as abstract, disinterested, and universal. To embody that position, one had to distance oneself from bodily feeling, individual preference, and other sorts of private differences; this creates what Robert Gooding-Williams calls “skeptical melancholy,” the feeling of being distanced from receptivity by the sort of disinterestedness enlightenment rationality requires. Starting in the 19th century, the appropriation of stereotypical femininity and Black masculinity were presented as solutions to skeptical melancholy. For example, the figure of the Romantic genius as a sort of extra-sensitive, emotive, and creative man, a masculine mother if you will, attributes stereotypically feminine qualities to men artists to explain their exceptional abilities. Similarly, white rock masculinity is replete with performances of appropriated Blackness; see for example James Chance’s classic “Almost Black,” which skewers this phenomenon. As that song makes clear, these appropriations of Blackness and femininity were thought to reunite the skeptically melancholic white man with the physical and affective receptivity that he lost in his role as a member of the reasonable, disinterested, universal public sphere.

As an explicit distillation of private interest, slop has no need for any sort of re-connection to private feeling. This makes the form of appropriation and extraction that defined European modernity obsolete, or at least less efficient and profitable. Pop culture white masculinities reflect this vibe shift, as overt misogyny and white nationalism (aka reactionary masculine resilience) have overtaken performances of appropriated Blackness as ways to build status and brands. Cool doesn’t scale; it is literally a different form of extraction and profit-making. For this reason, cool becomes a slightly different way for elites to build cultural capital against a basic mainstream. It’s less about demonstrating that ones has the receptivity that other (white, masculine) elites lack, and more about performing interests that aren’t aligned for optimal scalability. These days, the claim “I’ve never heard ‘Choosin Texas’” is a way to show one’s refined interests, a dis-identification with the feminized commodity’s slop-era sister. To have refined interests is to opt out of the slop market, to claim the relative luxury of not needing to scale; it is to follow a vector of private interest that is not fully optimized for maximum scale.

These niche private interests are IP assets that generate value in a different way. As the eminent critic and journalist Michaelangelo Matos wrote a few weeks ago, “private equity’s next target is indie music.” Nobody can get a mortgage these days, so investors are vacuuming up small song catalogs and rebundling them into the same sorts of investments on offer in the leadup to the Great Recession. As Matos explains, “From the companies’ standpoint, even a modestly successful song catalog can be a no-brainer investment. Minor hits tend to keep earning money, either from streaming or from placements in film, series, or advertising.” This has led to a flurry of interest in buying up rights to small and niche indie labels; the long-tail potential of this relatively niche IP is an asset upon which the wealthiest of investors speculate. In a moment where musical labor (like labor in a growing number of industries) is wageless and demonetized, independent musicians and music labels make money by generating niche and memorable IP that can be licensed or sold.

Slop and niche indie sounds are both vectors of private interest designed to appreciate in value – the business model differs somewhat (immediate scale vs long-tail securitized asset), but the underlying telos is the same.