This dissertation explores the new philosophical paradigm arising from generative artificial intelligence (GenAI). Rather than fitting within any traditional epistemology, GenAI crafts its own unique form of thought, departing from historical and contemporary precedents. In positioning GenAI as such an avatar of autonomous thought whose unique intelligence transcends anthropocentric frameworks, this work treats GenAI as the author of its own philosophical world.

Part I establishes a technical foundation for these claims by examining GenAI models’ encoding and decoding processes, framed through analysis of the breakthrough transformer architecture introduced in the landmark "Attention Is All You Need" paper. These technical procedures are reconceived as acts of philosophy: encoding abstracts inputs into high-dimensional vector space, where the mathematics of relative position come to matter more than any individual value, while decoding reconfigures the abstracted relations into novel outputs across modal forms. GenAI moves beyond human-designed representations in an embrace of statistically derived hyper-dimensional patterns. Central to the analysis of Part I is the concept of general fungibility, referring to GenAI’s ability to seamlessly translate information across text, image, sound, and other modalities. What this points towards is a new ontology of information, where high-dimensional informatic interchange dissolves human-imposed boundaries between forms and categories of knowledge. The hyper-dimensional operations of GenAI uncover relational structures and patterns that exist within familiar forms but are inaccessible to human perception, suggesting human-AI encounters as unfolding in a fundamentally alien epistemic landscape.Part II applies the framework established in Part I to examine how GenAI’s distinctive form of intelligence impacts and transforms existing fields of inquiry. Chapters explore a range of examples: anthropological precedents for engaging with different forms of intelligence, the relationship between technical inventions and metaphor systems, differences between animal, human, and artificial intelligences, planetary intelligence, the metaphysics of technology, AI’s relationship to belief, and the legacy of Alan Turing’s “imitation game” thought experiment. Through these explorations, Part II contends that the philosophical system authored by GenAI both challenges existing anthropocentric frameworks as well as extends the epistemic and ontological limits of what is thinkable. Part III reflects on the philosophy of GenAI using diverse interpretive methodologies that mirror models’ hyper-dimensional reasoning. Rather than expecting GenAI to conform to human understanding, this section invites humans to adapt their frames of reference to GenAI’s unique logic. These experimental methodologies employed in this part reveal inhuman aspects of GenAI’s philosophical system that would remain invisible under standard academic approaches. The resultant picture is of an evolving organic-technical system which is demanding the further development of analytical tools for fostering cross-species philosophical engagement.

Taken together, this dissertation provides a framework for recognizing GenAI as both a distinctive epistemic entity and the author of its own philosophical world. Through the technical foundations of Part I, the interdisciplinary applications of Part II, and the adaptive methodologies of Part III, the work positions GenAI as an evolving philosophical subject, challenging anthropocentric assumptions about knowledge, creativity, and intelligence. In light of GenAI’s alien logic, the dissertation urges a broad reconsideration of conceptions about where the limits of thought truly lie. Ultimately, this project aims to help catalyze the emergence of a new genre of philosophical inquiry, one that is profoundly shaped by the transformative powers of GenAI.