Artificial Intelligence in Science and R&D: Applications and Implications (ScientIA)

ScientIA, a 48-month project funded by the ANR as part of the 2021 generic call for projects, is coordinated by Floriana Gargiulo (Gemass UMR 8598, CNRS and Sorbonne Université) with Tommaso Venturini (CIS UPR 2000, CNRS), Jean-Gabriel Ganascia (LIP6 UMR 7606, CNRS and Sorbonne Université) and Vittorio Loreto (Sony CSL).

Associated with the deluge of big data, Artificial Intelligence (AI) has gradually moved out of the fields where it sprung from – computer science and mathematics – to spread to all areas of scientific and industrial research.

Introduced as an exogenous research technology, AI reshaped the scientific “adjacent possible”, extending the set of potential discoveries and innovations that are reachable within the available knowledge boundaries. This sudden expansion of the epistemic horizon produced, in the last decade, major advances in several research areas.

Because science is a complex interconnected system, one cannot fully understand the epistemic impact of AI without focusing on the related transformations of the scientific practices (working practices, politics of research, etc.) and the cognitive schemes, along with the unavoidable associated controversies.

This project stands on a robust interdisciplinary framework combining computational tools (network science, complex systems methods, AI techniques, modelling, etc.) and social science methodologies (surveys and qualitative interviews). ScientIA aims to analyze, with a comparative perspective, the impact and the implications of the introduction of AI in academic and industrial research. We will study the introduction of AI in these ecosystems of applied or basic research, with a multifaceted approach linking together the epistemic transformations that AI induced, the epistemological fractures it generated, the related renewal of the scientific practices, and the large-scale societal perception. 

The project revolves around three questions:

  • (Q1) What mechanisms favoured the spreading of AI in different research areas, and what feedback had this spreading on AI?
  • (Q2) Has the entrance of AI in science and R&D had significant consequences on scientific cultures and scientific practices?
  • (Q3) What is the social acceptance of the ever-increasing use of AI in science and technology?

These research questions are addressed through four scientific working packages having the following objectives:

  • (WP2) Identifying entrance patterns of AI in the selected case studies, going from identifying the diffusion drivers to the in-depth analysis of the most important AI tools mobilized and created in each case study (Q1).
  • (WP3) Exploring the impact of AI on epistemic innovation in the scientific ecosystem and predicting its future evolution (Q2).
  • (WP4) Exploring the impact of AI on the scientific practices and epistemological structures (Q2).
  • (WP5) Analysing AI technologies’ public image of its evolution (Q3).

The interdisciplinary consortium of ScientIA will develop an innovative research framework on AI, filling the large gap between scientometric studies, the growth of AI-based scientific research, and the general analyses of AI’s impact on society.

Starting from the analysis of the innovation patterns related to the introduction of AI in research, ScientIA is, more generally, an original contribution to understanding the transformations of science and technology in the era of the digital revolution.

The outcomes of this project will also bring significant benefits for the industrial sector, showing which research investment in AI technology will be needed to sync with the market of future innovation.

This project is funded by the ANR.