Science in the Digital Age

Abstract

In contemporary philosophy of science and technology, the rapid development of artificial intelligence (AI) technologies brings to the forefront fundamental questions about the nature of scientific cognition and its transformation under the influence of algorithmic systems. A review of philosophical discussions on intelligent systems in science demonstrates that over three decades there has been a fundamental shift in the point of consensus: whereas at the initial stage they were considered exclusively as a tool for enhancing human cognitive abilities, in 2022–2023 the philosophical community found itself facing the need to legitimize AI as a nontrivial epistemic participant. The traditional question of whether AI can be regarded as a cognitive agent has been supplemented by problems related to the practices of its use: under the influence of big data and neural networks, the criteria of scientificity are being transformed, the subject – object distinction is facing a crisis in the face of algorithmic opacity, and new models of epistemic responsibility are being sought. AI has established itself as a functionally irreducible participant in cognitive practices; however, none of the proposed conceptions – from extended instrumentalism to distributed or collaborative agency – completely eliminates the conceptual tension created by the gap between AI’s effective performance of cognitive operations and its lack of intentional states. This makes the direct application of traditional criteria of justification impossible. At the same time, the main discussions remain within the framework of first-order epistemology, focused on the question of the (quasi-)subjectivity of machines. However, this research program, for all its productivity, overlooks a fundamental transformation: AI's acquisition of a metascientific role, in which it not only actively participates in cognition but also largely restructures the configuration of the research field.