华东师范大学学报(哲学社会科学版) ›› 2026, Vol. 58 ›› Issue (4): 1-6.doi: 10.16382/j.cnki.1000-5579.2026.04.001

• 重启科技与人文对话 •    

人工智能为何没有意识

翟振明   

  • 接受日期:2026-06-17 出版日期:2026-07-15 发布日期:2026-08-03
  • 作者简介:翟振明,中山大学哲学系教授(广州,510275)

Why Artificial Intelligence Lacks Consciousness

Zhenming Zhai   

  • Accepted:2026-06-17 Online:2026-07-15 Published:2026-08-03

摘要:

人工智能是否具有意识,可以从意识本质、人工智能运作机制、意识不可复制的依据及人工智能定位四个维度层层递进。首先,意识具有主观体验与生物基础双重核心:主观体验层面,“感受质”和“玛丽的房间”实验解释了意识的私密性与不可传递性;生物基础层面,脑区功能对应关系与临床案例表明,意识是大脑神经元生物过程的涌现产物,并可区分为“现象意识”与“访问意识”以揭示意识的本质与功能属性。其次,人工智能的运作本质是“大数据训练+算法优化”模式下的符号操作与功能模拟,存在无主观体验、无自我意识、无创造性突破三大局限,“中文房间”实验与“哲学僵尸”概念进一步佐证人工智能与人类意识的本质差异。再者,从哲学思想实验(中文房间、哲学僵尸、意识替换)与神经科学机制(脑区功能特异性、量子过程潜在作用、神经可塑性)双重维度可以证明,意识依赖主观体验与生物基础,具有不可复制性。基于此,人工智能应被定位为人类能力的“工具延伸”,其价值在于增强人类的数据处理与机械劳动能力,而人类意识所具有的主观体验、意义建构与道德责任能力则具有不可替代性,未来应秉持“人机协作”模式,实现人与技术的和谐共生。

关键词: 人工智能, 意识, 感受质, 中文房间, 神经现象意识

Abstract:

The question of whether artificial intelligence (AI) possesses consciousness can be examined through four progressively unfolding dimensions: the nature of consciousness, the operational mechanisms of AI, the basis for the non-replicability of consciousness, and the positioning of AI. First, consciousness has the dual cores of subjective experience and biological foundation: At the level of subjective experience, “qualia” and the “Mary’s Room” experiment explain the privacy and non-transmissibility of consciousness; while at the level of biological foundation, the correspondence between brain regions and functions, as well as clinical cases, demonstrates that consciousness is an emergent product of the biological processes of brain neurons, and it can be distinguished into “phenomenal consciousness” and “access consciousness” to reveal its qualitative and functional attributes. Second, the operational essence of AI is symbol manipulation and functional simulation under the “large-scale data training + algorithm optimization” model, which has three major limitations, that is, the absence of subjective experience, the absence of self-awareness, and the absence of genuine creative breakthroughs. The experiment of “Chinese Room” and the concept of “philosophical zombie” also corroborate the essential difference between AI and human consciousness. Furthermore, through the dual perspectives of philosophical thought experiments, including “Chinese Room”, “philosophical zombie”, and “consciousness replacement”, and neuroscientific mechanisms, including the functional specificity of brain regions, the potential role of quantum processes, and neuroplasticity, it can be proved that consciousness is non-replicable for it depends on subjective experience and a biological basis. On these grounds, AI should be positioned as a “tool extension” of human capabilities, with its value lying in enhancing human abilities in data processing and mechanical labor, while the subjective experience, meaning-making, and moral responsibility of human consciousness remain irreplaceable. In the future, a “human-machine collaboration” model should be upheld to achieve harmonious coexistence between humans and technology.

Key words: Artificial Intelligence (AI), consciousness, qualia, Chinese Room, neural phenomenal consciousness