The Return of Subjectivity: the Second Enlightenment from Turing’s Thought

This article argues that by proposing the Turing Test as a criterion for determining whether a machine possesses intelligence, Turing fully embraced human subjectivity. Turing previously introduced the concept of the oracle Turing machine model, whose practical significance lies in providing an open architecture for dynamic computation involving “humans within the system”.Although no computer has achieved “divine” capabilities, we have witnessed the emergence of computational systems that are “human-centric”—exemplified by the Internet. The services provided by this system have transcended the scope of classical Turing-computable functions, enabling every individual to witness the immense power generated by the stimulation and convergence of collective human intelligence via the Internet. Unintentionally, this has also laid the data foundation for big data-driven machine learning, catalyzing the birth of large language models (LLMs) and propelling humanity’s informatization process into an advanced stage of intelligence. This article contends that the practical significance of the success of large language models lies in humanity’s entry into a new era of human-machine collaboration achieved through natural language, marking a new height in human subjectivity. At this critical juncture, humans should neither doubt nor lose their sense of subjectivity.