回应与承认从黑格尔意识经验看人工智能与自我意识问题

Abstract

The question whether generative artificial intelligence is conscious is usually framed as a dispute between functionalism and the theory of inner experience. Rather than following this path of external adjudication, this paper returns to Hegel's account of the "experience of consciousness" and his theory of recognition in the Phenomenology of Spirit, and reformulates the question: not "does artificial intelligence possess consciousness?" but "can artificial intelligence be recognized as another self-consciousness?" The object of inquiry is restricted to contemporary generative artificial intelligence represented by large language models. This paper argues that what matters is not which features of consciousness artificial intelligence has, but what it lacks. Current generative artificial intelligence does not satisfy the conditions for occupying the position of another self-consciousness that could perform the function of recognition: it cannot undergo Hegel's experience on its own and cannot accomplish self-negation; it lacks "meaning-as-such" (Meinung), having never passed through the cutting of the symbolic order; it may reach the level of understanding in appearance, but it cannot ascend beyond understanding, and its features of understanding are conferred from without by human beings. Its movement of trying to overcome external determination is merely a "dialectical simulacrum" or a "self-negation without a subject." Hence it can mediate the human relation to itself, but it cannot occupy the place of another self-consciousness. The danger of the digital age lies not in artificial intelligence actually becoming a subject that recognizes me, but in our mistaking a linguistic response at the level of understanding for mutual recognition. Key words: Hegel; experience of consciousness; recognition; artificial intelligence; self-negation; dialectical simulacrum