Generative AI: Remix or Slop?

Is generative AI simply another chapter in the long history of cultural remix, or has it crossed a threshold into something closer to pollution?

The accusation that AI merely “remixes” its training data is now familiar. Critics argue that these systems do little more than recombine patterns drawn from vast cultural archives, producing outputs that appear derivative, even parasitic. Yet this charge rests on a particular understanding of creativity — one that begins to wobble the moment we look more closely at how culture has always worked.

Two influential framings sharpen this tension. In the video-essay Everything is a Remix, Kirby Ferguson argues that creativity is built from copying, transforming, and combining. Hip hop sampling, rock riffs, memes, cinema, even the architecture of the web itself all testify to a culture that advances through recombination. On this account, derivation is not a flaw but the engine of innovation. If everything is a remix, then generative AI’s recombinatory logic may be less revolutionary than we imagine.  

By contrast, Kate Crawford’s essay Eating the Future: The Metabolic Logic of AI Slop introduces a more unsettling metaphor. Generative AI’s outputs, she suggests, function as “slop” — mass-produced, low-nutrient content designed for efficiency rather than nourishment. Framed metabolically rather than genealogically, AI becomes part of an extractive system that digests cultural material and produces endless output with little regard for its qualitative or imaginative consequences. The risk is not simply derivative art, but a malnourished cultural environment.

The Shared Question

Placed side by side, these two texts can be read in productive tension. Both agree, implicitly, that culture is derivative. Where they diverge is in how they evaluate that condition. Ferguson appears to celebrate it: derivation is creativity’s lifeblood. Crawford warns against it: at industrial scale, derivation becomes degradation. Her emphasis on metabolism cautions against romanticizing copying without attending to scale, power, or waste. Ferguson’s interest in genealogy, by contrast, cautions against pathologizing derivation without recognizing the long tradition of remix as the engine of creativity.  

Reading them together crystallizes an important question: when does copying nourish culture, and when does it begin to pollute it? This question is more than aesthetic. It cuts to how we understand imagination in the machinic age. Is imagination a chain of borrowings that link us across time (lineage), or is it a system of circulation that can be overfed, malnourished, or poisoned by the logics of technology and capital (metabolism)?  

Three insights emerge from placing these texts in dialogue.

1. Thresholds of value
   Not all remix is equal. While hip hop’s sampling of James Brown created new worlds of sound, the flood of AI-generated images can sometimes feel more like churn than innovation. Ferguson’s genealogy shows that copying is always the starting point. Crawford’s critique reminds us that not all copying leads to sustenance. Somewhere between remix and slop lies a threshold we need sharper tools to recognize and articulate.
2. The politics of who remixes
   Ferguson’s story centres on artists: DJs in the Bronx, programmers tinkering with code, filmmakers drawing on genre. His remix is democratic, even utopian. Crawford shifts the lens to institutions: the corporations building trillion-parameter models by scraping the world’s cultural output without consent. The difference is not just scale but politics. A Bronx DJ looping records is not the same as a tech giant feeding the world’s art into a machine. The question is not only what gets remixed, but who has the power to remix, and to what end.
3. The metabolism of imagination
   The most generative tension lies in the metaphors themselves. “Remix” is genealogical: it highlights inheritance, connection, and the lineage of ideas. “Slop” is metabolic: it highlights digestion, waste, the circulation of cultural nutrients and toxins. Read together, they remind us that imagination is both. It is lineage and metabolism, continuity and circulation. What matters is whether the system feeds or starves our capacity to imagine. 

Concluding thoughts

These texts matter in the context of ongoing debates regarding AI and the imagination. As provocations, they surface the foundational ambiguity of AI’s creative role and provide a sharpened sense of the stakes. On one hand, AI can be understood as just another remix machine, extending the copy-transform-combine logic that has always structured creativity. On the other hand, its industrial scale, extractive training methods, and endless generative churn suggest something more troubling: a cultural metabolism tipping into pollution.

This ambiguity is not something to resolve once and for all. It is a generative problem — a way of framing AI’s relation to imagination that resists both uncritical celebration and outright dismissal. From different angles, Crawford and Ferguson press the same underlying question that opens onto others:

• Where do we draw the line between homage, tradition, and theft?
• How do scale and power change the meaning of copying?
• Can we cultivate practices of remix that are nourishing rather than polluting?

This is fertile ground for thinking about AI and imagination — not as a debate over whether machines can be creative, but as an inquiry into the conditions under which creativity feeds us, and the conditions under which it starves us.