Framing the effects of machine learning on science - Research framework and emerging issues

Aug 5, 2026

Upload by Hazel

Studies on artificial intelligence and science often take a partial perspective, leaving the pathways of influence insufficiently synthesized. This work systematically maps how recent machine learning techniques, including deep learning, affect science using a taxonomy derived from Nathan Rosenberg’s effects of technology on scientific development. Four categories are proposed: intellectual, experimental, economic, and instrumental. Applying the framework identifies multiple triggers in scientific practice and highlights two urgent issues: experimental effects concentrate in a few companies, and new techniques diffuse without explanation across scientific disciplines.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages17

Reading Time43 min

AccessFree to download

Introduction

Rosenberg Effects framework

Applying the framework to ML/DL

AI & SOCIETY (2024) 39:749–765 [https://doi.org/10.1007/s00146-022-01515-x](https://doi.org/10.1007/s00146-022-01515-x) Framing the effects of machine learning on science Victo J. Silva1,2 · Maria Beatriz M. Bonacelli1 · Carlos A. Pacheco1 Received: 17 January 2022 / Accepted: 25 May 2022 / Published online: 24 June 2022 © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2022 Abstract Studies investigating the relationship between artificial intelligence (AI) and science tend to adopt a partial view. There is no broad and holistic view that synthesizes the channels through which this interaction occurs. Our goal is to systematically map the influence of the latest AI techniques (machine learning, ML and its sub-category, deep learning, DL) on science. We draw on the work of Nathan Rosenberg to develop a taxonomy of the effects of technology on science. The proposed framework comprises four categories of technology effects on science: intellectual, economic, experimental and instrumental. The application of the framework in the relationship between ML/DL and science allowed the identification of multiple triggers activated by the new techniques in the scientific field. Visualizing these different channels of influence allows us to identify two pressing, emerging issues. The first is the concentration of experimental effects in a few companies, which indicates a reinforcement effect between more data on the phenomenon (experimental effects) and more capacity to commercialize the technique (economic effects). The second is the diffusion of new techniques lacking in explanation (intellectual effect) throughout the fabric of science (instrumental effects). The value of this article is twofold. First, it provides a simple framework to assess the relations between technology and science. Second, it provides this broad and holistic view of the influence of new AI techniques on science. More specifically, the article details the channels through which this relationship occurs, the nature of these channels and the loci in which the potential effects on science unfolds. Keywords Artificial intelligence · Science and technology interaction · Nathan Rosenberg · Deep learning · Intellectual debt 1 Introduction Besides becoming one of the central technologies of the alleged fourth industrial revolution (Yu et al. 2021), artificial intelligence-based systems influence more than commercial trends. They impact diverse social spheres and their influence extends to scientific research (Chubb et al. 2021) . The scholar community recognizes that “Scientific research can lead to technological advance, but technology very much affects advances in science" (Stephan 2010, p. 229) . The same applies to recent artificial intelligence (AI) * Victo J. Silva [victont@gmail.com](victont@gmail.com) 1 Geosciences Institute, Universidade Estadual de Campinas (University of Campinas), Carlos Gomes st., 250, Campinas, São Paulo 13083-855, Brazil 2 Copernicus Institute for Sustainable Development, Utrecht University, Vening Meinesz Building, Princetonlaan 8a, Utrecht 3584 CB, The Netherlands techniques, such as machine learning (ML) and deep learning (DL) . Still, available studies offer a partial account of their influence on science. Cockburn et al. (2018) investigate bibliometric data and patents. They conclude that there was a reorientation toward applied deep learning solutions from the 2010s onwards. More than that, they conclude that DL systems alter knowledge-producing circles such as science. Vasilescu and Filzmoser (2021) conceptualize machine invention systems, which share a similar view on the invention in the method of invention (IMI) . Bianchini et al. (2020) are more specific: they investigate how different areas of research instrumentalized DL systems. They demonstrate its use as a tool and provide insights into the effects of this application for the health sciences research

FAQs

What is the article’s main goal regarding AI and science?

To provide a systematic, holistic mapping of how recent machine learning techniques influence science through identifiable channels rather than isolated partial studies.

What are the four categories of technology effects on science in the proposed framework?

The framework groups effects into intellectual, experimental, economic, and instrumental categories, derived from Rosenberg’s taxonomy.

What two emerging issues does the analysis of ML/DL reveal?

Experimental effects are concentrated in a few companies, suggesting reinforcement between data-driven experimental capacity and economic commercialization, and new techniques spread through science while lacking explanatory grounding.

Aug 5, 2026

Upload by Hazel