arXiv cs.CY· Sasha Luccioni, Travis LaCroix·· 11 小时前AI 评分31
论文探讨“F*ck AI”口号背后的道德恐慌与“AI”语义不确定性
What Do We Mean When We Say 'F*ck AI'?: Moral Panic, Semantic Indeterminacy, and the Reification of Techno-Capitalism
AI 导读
一篇 arXiv 论文指出，“AI”作为统称缺乏单一明确所指，围绕它的道德恐慌反而把 AI 固化为一个自主、单一的智能体形象，与企业叙事相吻合。作者认为，善意的批评与公众抗议因此 paradoxically 放大了 AI 炒作，并让针对“AI”的治理政策必然失效。论文最后提出提升 AI 素养、强化问责和改造现有监管的路径。
正文 · 原文
Abstract:Academic and (increasingly) public resistance to artificial intelligence (AI) often blurs the line between legitimate critique of techno-capitalist exploitation and (inflated) existential risk. Moreover, discourse around AI can sometimes resemble a "moral panic" where an apparently clear and identifiable scapegoat is targeted in lieu of the complex systemic problems that are causing (genuine and justified) social anxieties. However, AI is not a monolith. In this paper, we explore the environmental, political, and socio-technical consequences of the semantic indeterminacy of discourse about "AI" in conjunction with the increasing moral panic encompassing these technologies. We argue that since the umbrella term "AI" lacks a coherent, singular referent, the moral panic about AI inadvertently reifies the concept of an autonomous, singular agent with inherent and often autonomous tendencies, which aligns with corporate narratives. Hence, well-intentioned critique, activist pushback, and public protest paradoxically amplifies hype-based discussions of AI, thus validating the "threat" narrative that tech leaders deploy to exaggerate their products' capabilities and market value, further consolidating and centralising their political and socio-economic power. Moreover, if "AI" does not refer to a unified class of technologies, then policy intended to govern "AI" will systematically (and necessarily) fail. We conclude by proposing ways forward to improve AI literacy, enforce accountability, and repurpose existing regulation in order to imagine a more positive AI future.
| Subjects: | Computers and Society (cs.CY) |
| Cite as: | arXiv:2611.05574 [cs.CY] |
| (or arXiv:2611.05574v1 [cs.CY] for this version) | |
| https://doi.org/10.48550/arXiv.2611.05574 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Alexandra Sasha Luccioni [view email]
[v1]
Tue, 3 Nov 2026 20:06:08 UTC (3,763 KB)
来源：arXiv cs.CY · arxiv.org