Artificial intelligence grounded in the comprehensive, syntagmatic, and holistic understanding of science... and vice versa

Abstract

The rapid expansion of artificial intelligence (AI) tools for academic research has generated an ecosystem of resources that, despite their technical diversity, share a common characteristic: the absence of an explicit epistemological framework to guide their suggestions and outputs. This article describes the experience of designing Tutoeris, an AI platform built to assist researchers in training throughout the development of their degree projects, constructed on the basis of the holistic understanding of science formulated by Jacqueline Hurtado de Barrera and Marcos Fidel Barrera Morales. It is argued that the holistic understanding of science can function not only as a theoretical framework for thinking about AI, but as a functional architecture for designing it: its principles translate into generation rules, its concepts into structured forms, and its spiral of the research process into a coherent sequence of steps for the researcher. At the same time, the inverse direction of the relationship is examined: what AI contributes to the holistic research process and what its real limits are. The article concludes that epistemologically informed AI (that which is designed from a specific methodological framework rather than from the apparent neutrality of generalist models) represents a necessary and viable direction for the development of technology in the service of science.