Galip, I., & Lupinacci, L. (2026). Degenerative AI: AI slop, brainrot, and the gimmick of generative culture. European Journal of Cultural Studies. https://doi.org/10.1177/13675494261468638
Summary
This short theoretical essay examines the generative-AI content and practices popularly derided as “slop” and “brainrot,” treating them simultaneously as an aesthetic genre of (often automated) digital content and as affective modes of being and feeling bound up in their consumption. Drawing on Sianne Ngai’s Theory of the Gimmick, Galip and Lupinacci argue that these stigmatised forms operate as “overrated devices” that work both “too little” — through labour-saving tricks and deskilling — and “too hard” — through strained algorithmic efforts to seize attention. The result is content that is “always enchanting and repulsive at once.” The authors contend that slop and brainrot lay bare the contradictions of cultural production under “too late capitalism,” while probing whether such ostensibly degenerative forms retain any critical or resistant potential.
Key Contributions
- Extends Ngai’s aesthetic category of the gimmick to generative-AI content, providing a lens for the mix of captivation and dissatisfaction it provokes.
- Offers a theoretical vocabulary — labour/value/time, deficiency/excess, “frictive forms” — for cultural-studies analysis of synthetic content.
- Reframes “degenerative” and “unproductive” genres as objects worthy of serious critical inquiry rather than dismissal.
- Connects slop and brainrot to broader debates on platform “enshittification,” authenticity, and aesthetic sanitisation.
Methods
A conceptual and theoretical essay rather than an empirical study — no datasets are generated or analysed. The authors apply Ngai’s Theory of the Gimmick (2020), organised around three axes of labour, value, and time, alongside a discourse analysis of contemporary public commentary, journalism, tech-industry rhetoric, and emerging scholarship. Illustrative cases include “Italian Brainrot” (e.g. Tralalero Tralala, Ballerina Cappuccina), the AI microdrama Fruit Love Island, and platform products such as Meta’s Vibes and OpenAI’s Sora.
Findings
- Labour: slop is marketed as “effortless” but decried as “craftless” — the ultimate labour-saving gimmick that alienates creators from their output while still relying on outsourced low-wage labour and new cleanup work (fixing botched outputs, prompt engineering).
- Value: slop is figured as flagrantly worthless “digital trash,” likened to ultra-processed food, yet depends on enormous financial and environmental costs; “slop farms” largely fail to monetise, with creators profiting mainly from selling “how-to-slop” tutorials.
- Time: generative AI is ahistorical and decontextualised, sustained by hype-driven temporal manipulation (“it is already too late”), yet recursive — systems degrade when fed their own outputs (“Model Autophagy Disorder”), so the machines’ own “brains” can be rotted.
- Brainrot is discursively split between innocent nonsense and deceptive material with hidden messages, and between a cognitively eroding effect and a decompression-based genre of participation (“anaesthetic media”) valued precisely for being unproductive.
- As gimmicky “sad marvels,” slop and brainrot combine sophisticated computation with perceived impoverished, deskilled outputs — enchanting and repulsive at once.
Connections
This essay sits within the imaginaries strand of the register, offering a cultural-theory counterpoint to the more descriptive and computational accounts of AI-generated media. It resonates most directly with work theorising synthetic culture and aesthetics, such as Manovich2026-ih, and complements empirical mappings of AI slop and manipulated media like Baym2026-tr and Hepp2026-oi; the discussion of platform products and content ecosystems also bears on infrastructural analyses such as Helmond2026-ll.
Podcast
A research-radio episode discusses this paper: 🎧 MP3 · Spotify