<p>This article argues that prompt engineering, the practice of writing textual inputs to shape AI outputs, resembles the psychological and temporal structures that Walter Benjamin identified in gambling behaviour. Users of large language models have to work with a measurably aleatory process as they phrase instructions, engineer context, adjust tone, and iterate across repeated runs, only to find that identical inputs produce different outputs and minor wording changes cascade through the probability field of the generated text. I call this condition “prompt anxiety” to describe a key feature of how stochastic systems organise cognitive labour under what I call vector capitalism. To ground this claim, I have developed LLMbench, a research instrument for the comparative close reading of LLM outputs that visualises token probability distributions, entropy curves, and cross-model divergence. Analysis through LLMbench demonstrates that the uncertainty users experience corresponds to measurable variation in model confidence across the generated text. Drawing on Benjamin’s concept of contingency alongside Marx’s analysis of labour and Gramsci’s theory of hegemony, the article traces how AI platforms transform this uncertainty into extractable value through subscription models, token-based pricing, and prompt marketplaces. The collective practices that emerge in response, from shared prompt strategies to jailbreaking techniques, represent vernacular knowledge formations that, whilst often exhibiting magical thinking, contain resources for what I call “revolutionary prompting” and the transformation of individual prompt anxiety into collective political critique of the conditions of AI production.</p>