How reality turns to slop — the AI feedback loop
AI-generated content shapes culture by shifting what seems normal, then that shifted culture feeds back into AI training data, creating a cycle toward “hyperslopification”.
The mechanism: AI slop is hyperpalatable (easy to consume, optimized for engagement), so it crowds out ordinary content. Over time, the training distribution drifts toward the weird and cartoonish, and models learn from a warped base reality.
Real examples cited include memetic drift in social media, image synthesis poisoning downstream datasets, and the cultural homogenization effect as slop becomes default.