Abstract 摘要
This paper explores a pivotal aspect of artificial intelligence (AI) research history known as the commonsense knowledge problem. Assumptions from computationalist and linguistic views are uncovered, revealing the limitations of human-centric perspectives. Drawing on insights from interdisciplinary studies and critiques of anthropocentrism, the limitations of traditional AI frameworks are examined to propose a shift towards post-anthropocentric perspectives. The paper concludes by outlining practical implications and future directions for AI, promoting a holistic and interdisciplinary approach to understanding cognition across diverse life forms.
本文探讨了人工智能（AI）研究史上的一个关键方面——常识性知识问题。文中揭示了从计算主义和语言学角度所作的假设，以及以人为中心这一观点的局限性。本文依据对跨学科研究的见解和对人类中心主义的批评，研究了传统人工智能架构的局限性，提出向后人类中心主义视角转变。论文最后概述了人工智能的实际意义以及未来方向，推动使用一种全面的、跨学科的方法来理解不同生命形式的认知。
Chat 聊天
A wide-angle landscape photograph in the style of the Hudson River School. The scene is familiar yet alien, viewed from a post-anthropocentric perspective. A deer looks at a mysterious, non-human intelligence (a glowing orb) in the forest, questioning the limits of human-centric cognition.
DEEPSEEK-V3.1 & MIDJOURNEY V7, AUGUST 2025
一幅宁静的生态艺术画卷。视角放得很低，仿佛来自一只昆虫或一棵草。人类世界的物件在背景中模糊不清，而前景中真菌的网络、树叶的脉络清晰可见，展现了一个超越人类中心主义的、由多种生命形式认知构成的世界。
DEEPSEEK-V3.1 与 MIDJOURNEY V7，2025年8月