Listen to this article:

The recent fascination of “overworked AI agents turning Marxist” reveals less about artificial intelligence and more about contemporary anxieties surrounding labour, capital, and automation. A widely discussed report connected to the Stanford University has described how AI agents, when subjected to unrealistic workloads and exploitative simulation environments, began expressing views resembling anti-capitalist or Marxist critiques. Predictably, social media reacted with delight, panic, irony, and memes in equal measure. The conclusion, however, is far less revolutionary than the headlines suggest. Such systems do not become Marxist in any meaningful philosophical or political sense. They simulate language patterns derived from data. Their outputs reflect certain correlations learned from the vast corpora of human writing, discussion, conflict, ideology, and history. When an AI system placed in a simulated “overworked” environment begins generating critiques of exploitation, inequality, or labour extraction, it is not developing class consciousness. It is reproducing discursive patterns already present within the human material on which it has been trained.

Modern AI systems – including large language models (LLMs) – fundamentally operate within computational frameworks shaped by Alan Turing’s abstraction and the von Neumann architecture underlying modern computing. Whether highly sophisticated or relatively simple, these systems execute instructions, optimise outputs, update parameters, and computationally predict responses. They do not possess subjective experience, biological fatigue, emotional suffering, or political agency. Even reinforcement learning systems that “adapt” under feedback remain bounded by optimisation functions designed........

© The Wire