Yellow Coolant and Kellogg’s Diarrhea: Slop, Shanzhai and Cursed AI as Memetic Aesthetic Detournement

Abstract

The endless iterations of generative AI output oozing through digital social networks has become unofficially designated as “slop.” Often characterized as a bland averaging of training data or hallucinatory distortions, “slop” has come to symbolize dissatisfaction with an excess of synthetic media. While bland recombination hold the potential to become insidious tools of fascism, proffering a simple, clean hegemonic big-data aesthetic (Watkins, 2025; Salvagio, 2023), by focusing on practices of subversion and misuse, we argue for a reframing of generative AI images which sees a potential for detournement of the neoliberal socio-technical apparatus from which these images spawn. We analyze the Cursed AI Facebook group, a memecultural community fueled by a cadre of artists dedicated to misusing AI for nightmarish visions of intellectual property. Using theory from the mimetic re-turn and mimetic studies, we apply a Shanzhai framing to Cursed AI to show how mimetic processes can yield difference through iterative production and successive variation. In the context of Cursed AI, to engage in this participatory process of sharing, iteration and spread is to enter into the relational flux of memetic media and be $2 . We argue that the potential for rupture with AI-generated images comes, not from content or form, but through their potential to initiate and proliferate mimetic relational flux. This paper contributes to ongoing debates on digital aesthetics, AI-generated images, and the mimetic re-turn. It will be of interest to scholars of post-digital culture, aesthetics, meme studies, tactical and alternative media, and critical AI studies.