AI slop is a vegan hamburger

A viral post on the AI forum LessWrong, titled 'AI slop is a vegan hamburger,' uses a food analogy to critique the experience of consuming AI-generated content. Author 'pku,' drawing from experience in vegan social circles, compares tools like ChatGPT to products like the Impossible Burger: initially impressive in mimicking the real thing (human writing), but ultimately revealed as bland, uniform, and predictable upon sustained exposure. The core argument hinges on information theory, suggesting AI output has lower Kolmogorov complexity—it's easier for the brain to compress into a simple pattern, robbing it of the surprising richness found in human-created work. This predictability makes it feel grating, like 'someone wearing a bad mask of your ex wife’s face for breakfast every day.'

The post has ignited significant debate in the comments, with users dissecting the 'slop' label. One top comment by 'williawa' pushes back, suggesting the criticism may stem from a psychological bias akin to dismissing industrially produced goods or 'stuff made in china'—a reluctance to admit newer, cheaper methods can produce adequate results. The commenter notes that AI-written code, being more objectively evaluable, is already shedding the 'slop' label as it proves functionally effective, implying the same may happen for creative writing as models improve. The discussion highlights a central tension in AI adoption: distinguishing between genuine qualitative shortcomings and subjective resistance to disruptive, efficient new tools.