This article continues and develops the theme of the INION RAS Round Table "Science in the Digital Age," dedicated to the memory of V.I. Arshinov (February 26, 2026). The paper discusses problems and methodological foundations of a practice in which a large language model (LLM) serves as a working instrument of philosophical research and a node of the hybrid socio-techno-anthroposphere. Drawing on presentations by V.G. Budanov, M.F. Yanukovitch, and V.V. Tarasenko, a review of contemporary research on AI's impact on scientific inquiry, and the domestic program of machine epistemology, the author formulates four methodological functions of LLM applicable to philosophical work: topology of conceptual spaces, identification of "blind spots" in argumentation, generation of counterfactual axiomatics, and distributed corpus reading. These functions prove productive precisely because of LLM's structural constraints, primarily the impossibility of phenomenological epoche and the asymmetry of intentional structure: the semantic, arrative (⊙-dominant) mode of human thought and the computational, geometric (∇-dominant) mode of transformer operation form a bipolar configuration of boundary coupling. The concept of hybrid values emerging from human generative neural network interaction is examined. Values in this context form a new intentional structure and are defined as invariants of event interpretation at the boundary of heterogeneous systems. The concept of "hybrid scientific literature" is considered, along with practices of "corpus reading." The technology of "LLM-vision" is discussed, enabling the formation of a topological projection of a text corpus and adapting its complexity to the reader's cognitive capacities. In conclusion the principle of "episodic trust" for hybrid systems is proposed, based on limiting self-representation.
philosophy of artificial intelligence; machine epistemology; digital epoch; translational intentionality; intentional asymmetry; hybrid values; hybrid scientific literature; corpus reading; V.I. Arshinov's philosophy of complexity.