思维与意识的功能区分——从“反映—反应”模型看人工智能意识的边界

中文摘要

思维与意识的关系是心灵哲学、认知科学与人工智能哲学中的基础问题。围绕“机器是否有意识”“思维是否等同于体验”等争论，常见困难并不只在经验资料不足，也在于把对象表征、身体感受和反思判断混称为意识。本文提出一种功能性区分：思维是一阶反映，指对象状态在系统中的表征、调用、组合与推演；意识是主体反应，指主体系统对对象表征、身体状态和自身思维过程所发生的觉知、评价与调控。在意识内部，感受性意识以身体、内感受和情绪价值为基础，理性意识则是“思维对思维的反应”，表现为可报告、可反思、可修正的判断性觉知。由此，当前人工智能可被理解为具有复杂表征能力，并能在功能上模拟理性意识的某些方面，但不具有由生命身体支撑的感受性体验。该区分有助于澄清人工智能意识、机器理解和道德责任等问题

Abstract

The relation between thinking and consciousness is a foundational issue in philosophy of mind, cognitive science, and philosophy of artificial intelligence. Around disputes such as "whether machines are conscious" and "whether thinking is equivalent to experience," the common difficulties lie not only in insufficient empirical data, but also in conflating object representation, bodily feeling, and reflective judgment under the name of consciousness. This paper proposes a functional distinction: thinking is first-order reflection, referring to the representation, retrieval, combination, and inference of object states within a system; consciousness is the subject's response, referring to the awareness, evaluation, and regulation occurring in the subject system with respect to object representations, bodily states, and its own thinking processes. Within consciousness, phenomenal consciousness is grounded in the body, interoception, and emotional value, while rational consciousness is "thought's response to thought," manifesting as reportable, reflectable, and revisable judgmental awareness. Accordingly, current AI can be understood as possessing complex representational capacities and functionally simulating some aspects of rational consciousness, but it does not possess phenomenal experience supported by a living body. This distinction helps clarify issues of AI consciousness, machine understanding, and moral responsibility.

关键词：思维；意识；理性意识；现象意识；人工智能意识

Keywords: Thinking; consciousness; rational consciousness; phenomenal consciousness; artificial intelligence consciousness