Recent developments in artificial intelligence urge clarification of its ethical and legal status. The issue revolves around the concept of subjectness, distinguishing active and responsible conduct from inert performance. We analyze this notion from a physical viewpoint, building on the quantum-theoretic refinement of the concept of uncertainty into quantum and classical types: quantum uncertainty refers to an objective freedom to construct the future, while classical uncertainty denotes subjective ignorance of present states of nature. Subjectness of intelligence is then defined by the kind of uncertainty it is capable to resolve. To analyze different aspects of intelligence, quantum-inspired definitions of decision, subjectness, originality, and meaning are introduced on this basis. These concepts are first calibrated on natural intelligence and then applied to artificial systems, classified as classical and quantum. Classical AI then appears fundamentally alien to subjectness due to its algorithmic nature, limited to the resolution of classical uncertainty. Quantum AI, in contrast, breaks this limit by hosting some degree of proto-subjectness on the level of elementary particles, involved in its operation. Fundamentally, our approach tracks alternative views on subjectness of intelligence to the interpretations of quantum randomness, identifying both as different sides of the same ethical dilemma. Quantum physics then provides fertile ground for possible solutions, aligned with Eastern and Western views on freedom and constraint, subject and context in social life. These results offer a scientific approach to the controversial challenges of socio-technological development, integrating physical and humanitarian perspectives.

人工智能的最新发展迫切需要澄清其伦理和法律地位。这个问题围绕着主体性的概念，区分主动和负责任的行为与惰性表现。我们从物理角度分析这个概念，建立在量子理论将不确定性概念细化为量子和经典类型的基础上：量子不确定性是指构建未来的客观自由，而经典不确定性表示对当前自然状态的主观无知。情报的主观性是由它能够解决的不确定性类型来定义的。为了分析智能的不同方面，在此基础上引入了量子启发的决策、主体性、原创性和意义的定义。这些概念首先在自然智能上进行校准，然后应用于人工系统，分为经典系统和量子系统。由于其算法性质，经典人工智能似乎与主体性根本格格不入，仅限于经典不确定性的解决。相比之下，量子人工智能通过在参与其操作的基本粒子层面上承载一定程度的原始主体性，打破了这一限制。从根本上说，我们的方法追踪了关于智能主体性的不同观点与量子随机性的解释，将两者视为同一道德困境的不同方面。量子物理学为可能的解决方案提供了肥沃的土壤，与东西方关于社会生活中的自由与约束、主体和背景的观点保持一致。这些结果为解决社会技术发展的有争议的挑战提供了科学的方法，整合了自然和人道主义的观点。