Researchers propose new concept to study AI consciousness through subjective experience and mathematical theory
New research introduces an approach to studying artificial intelligence consciousness using phenomenology and mathematical category theory.
The AI academic community is seeing a new development with the publication of a research paper titled "Categorical AI phenomenology: A first-person approach," which proposes studying AI consciousness by using "phenomenology"—the study of subjective experience—as a starting point instead of merely looking at traditional physical mechanisms.
The paper is the work of Robert Prentner from the Institute of Humanities, ShanghaiTech University, and the Association for Mathematical Consciousness Science. It was published in the March 2026 issue of the Journal of Artificial Intelligence and Consciousness, with a preprint available on arXiv (arXiv:2608.20420).
The core of this research is defining consciousness as subjective experience that arises through an "interface" between an agent and the external world, utilizing mathematical tools from "category theory" alongside Q-network structures to create a mathematical model capturing phenomenological processes and structural behaviors of data processing systems.
This research marks a significant theoretical step forward in helping us understand the complexity of AI systems, going beyond traditional performance-focused developments.