Are AI Robots Human? – 智能機器人是人嗎？

Exploring the Confucian “Distinction Between Humans and Things” Through the Classical Confucian Mode of Naming “Things” – 從古代「物」之類別名談儒家的「人物之辨」

Abstract

When Confucian theorists examine the object of artificial intelligence—the robot (or su­perintelligence)—their terminology, shaped by conventions of naming in Chinese char­acters, frames it as an anthropocentric entity by designating it as ji qi (machine) + ren (hu­man), 機器人, rather than using the English term “robot.” In reflecting on human–robot ethics, they often employ the concept of the “distinction between humans and animals” (ren qin zhi bian 人禽之辨). The aim of this article is to clarify the classical Confucian framework for naming things that underlies the Chinese term ji qi ren, and to show that this designation does not imply that a robot is a human being, but rather that it resembles a human.

To avoid the cognitive bias embedded in the concept of ren qin zhi bian, which restricts discussion to “ethical evaluation” while neglecting the “ethical–descriptive” framework of classical Confucianism, this article analyzes how classical Confucians approached things in a pre-conceptual phase and how they classified and named different kinds of objects. In this way, it avoids the limitation of treating ji qi ren exclusively through the lens of human moral consciousness. Although Confucius addresses the naming of things from an anthropocentric perspective, his model of correct names incorporates both moral orientation and objective facts. By contrast, the treatments of the names of things by Zengzi, Mencius, and Xunzi are anthropocentric from the outset. Clarifying the classical Confucian framework for naming and classifying things according to their types and developmental processes can thus contribute to discussions of the distinction between hu­man beings and things (ren wu zhi bian 人物之辨) and enable the development of diverse perspectives on the ethical and cooperative relationships between them.

在面對人工智能（甚至有可能是超智慧形式物）時，多數的儒家研究者是以「人」作為主詞的「智能機器『人』」之「名」來討論，並且由「人禽之辨」的義理來思考人與該物的倫理方式。但是，此種思考模式僅有「倫理判斷式」（對人）而已，忽略了古代儒家所兼有的「倫理描述式」（對物）脈絡，這將無法與時俱進地產生有效的論述。準此，本文先梳理出博物階段的古代儒者是如何面對客觀存在「物」，補齊對「物」的脈絡，以此避免僅以道德意識、道德能動性的角度來看人與「類人」的課題，即：不僅是以價值範導性來看「物」，同時也以客觀真實性來看「物」。雖然孔子對待物名時，是「以人為始」(human-based) 的綜合關係，所以兼有價值範導性（興觀群怨）與客觀真實性（鳥獸草木）的「正名」，但是，直到曾子、孟子和荀子時，已然是在「以人為主」(human-cen­tred) 的宰治基礎上，而各自有「成人」與「成物」的論述偏向。通過釐清古代儒者對於「物」之類別名的概念架構來「類族辨物」，以此作為辨物基礎，再看「物」之類別名的前概念階段、概念秩序、概念從屬關係及其演變過程，然後從兩種論述偏向的儒者類型，即「成人」與「成物」視角來看「人物之辨」之下人與人工智能的關係，或可探索出不同視角的倫理思考方式。