Caroline Gans Combe

Caroline Gans Combe

Auto-generated profile, based on publicly available information

Caroline Gans Combe

INSEEC U. Institut National d’Etudes Economiques et Commerciales

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 February 2020 in Genome Medicine
Verified
Journal Article
Open access
PDF file
FAFrank Møller Aarestrup
Frank Møller Aarestrup
AAA. Albeyatti
A. Albeyatti
WAW. J. Armitage
W. J. Armitage
CACharles Auffray
Charles Auffray
LAL. Augello
L. Augello
HOH. Van Oyen
H. Van Oyen
The European Union (EU) initiative on the Digital Transformation of Health and Care (Digicare) aims to provide the conditions necessary for building a secure, flexible, and decentralized digital health infrastructure. Creating a European Health Research and Innovation Cloud (HRIC) within this environment should enable data sharing and analysis for health research across the EU, in compliance with data protection legislation while preserving the full trust of the participants. Such a HRIC should learn from and build on existing data infrastructures, integrate best practices, and focus on the concrete needs of the community in terms of technologies, governance, management, regulation, and ethics requirements. Here, we describe the vision and expected benefits of digital data sharing in health research activities and present a roadmap that fosters the opportunities while answering the challenges of implementing a HRIC. For this, we put forward five specific recommendations and action points to ensure that a European HRIC: i) is built on established standards and guidelines, providing cloud technologies through an open and decentralized infrastructure; ii) is developed and certified to the highest standards of interoperability and data security that can be trusted by all stakeholders; iii) is supported by a robust ethical and legal framework that is compliant with the EU General Data Protection Regulation (GDPR); iv) establishes a proper environment for the training of new generations of data and medical scientists; and v) stimulates research and innovation in transnational collaborations through public and private initiatives and partnerships funded by the EU through Horizon 2020 and Horizon Europe.

23 August 2023 in PLOS ONE
Verified
Journal Article
Open access
PDF file
CGCaroline Gans-Combe
Caroline Gans-Combe
JKJae Yun JUN KIM
Jae Yun JUN KIM
WMWaleed Mouhali
Waleed Mouhali
YRYves Rakotondratsimba
Yves Rakotondratsimba
ABAïcha Baccar
Aïcha Baccar
The notion of sobriety is considered a key variable in various energy transition scenarios. Often associated with a form of punitive ecology, it is, nevertheless, possible to make it a component that supports green growth, by linking it to the concept of "satisfaction". In this work, we have invented a way to achieve both “digital”, “economic”, and “ecological” sobriety, while ensuring the satisfaction of the end user. Directly correlated to the production of goods or services, the satisfaction function is built on the well-documented marginal utility function, which measures the need (or not) to consume further resources to satisfy the economic agents. Hence, it is justified and exists because it stands for the expectations of end users and makes sure the latter is met. This product itself is a function of the allocation of a set of resources, mapped using activity-based costing tools (ABC method). In this work, we focus on an AI proof-of-concept and demonstrate that it is possible to reach numerical sobriety by controlling the size of the training dataset while ensuring roughly the same model performance. In general, we show that it is possible to preserve the efficiency of AI processes while significantly minimizing the need for resources. In this sense, after establishing an analytical model, we suggest reducing the amount of data required to train the machine learning (ML) models, while guaranteeing zero change in terms of performance (say their accuracy). We show that it affects the energy consumed, and, thereby, the associated cost (i.e., economic and ecological) and the associated CO2eq emission. We thus confirm the existence of a "triangle of sobriety". It is defined as a virtual circle governed by a digital-economic-ecological sovereignty. We also propose that if AI production processes have a potential for sobriety, all identical activities have the same characteristics, thus opening the path to green growth.

2 April 2020
Verified
Book Chapter
Closed access
PDF file
CCCaroline Gans Combe
Caroline Gans Combe

5 May 2023 in European Conference on Social Media
Verified
Journal Article
Open access
PDF file
CCCaroline Gans Combe
Caroline Gans Combe
JJJae-Yun Jun
Jae-Yun Jun
WMWaleed MOUHALI
Waleed MOUHALI
Social media can help businesses and society implement sustainable practices (Bodin & Prell (Eds.), 2011). They also contribute to excessive resource use in data collection and sharing, digital resource use, and energy consumption (Kamin & Paireekreng, 2018). Thus, TikTok's feeds consume 15.81 mAh per minute and emit 2.63 gEqCO2/min in 2021. (Derudder, 2021). The generation that uses this tool most (Burns-Stanning, 2020) is also the most demanding and aware of climate change issues (Knight, 2016), even though its practices contribute to the current degradation. This hiatus illustrates how difficult it is to reconcile individual and collective goals. We propose shifting the question from the end-user to the service provider by detecting sobriety pits in data use and access to perpetuate or avoid disrupting end-user practices. Sobriety (frugality) is a critical variable in energy transition scenarios (Balzani, 2019) and one of the ecological transition's pillars. Social media and AI are particularly tense on these issues (Stein, 2020). Current solutions often pit energy efficiency against performance (Ikhlasse, Benjamin, Vincent & Hicham, 2021). Given the industry's promises of information access, automation, intelligent decision-making, human error avoidance, and more, this is unacceptable. Thus, we propose an operational framework to pose a model of sobriety for data access, consumption, and use (such as in social media or AI agents training) that does not impair performance or accuracy.

10 March 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
CCCaroline Gans Combe
Caroline Gans Combe
This paper argues that the rise of agentic AI—systems capable of autonomous perception, planning, and action—poses a structural threat to democratic life that conventional AI ethics frameworks fail to diagnose. Drawing on Hannah Arendt's analyses of totalitarianism, the vita activa, and the banality of evil, the paper identifies the deepest risk of agentic AI not as misalignment but as axiological displacement: the systematic transformation of what a society recognises as valuable through the progressive delegation of judgment to autonomous systems. The paper makes three contributions. First, it establishes an Arendtian diagnostic framework for agentic AI, showing that these systems reproduce the cognitive conditions Arendt identified as precursors to democratic collapse—not through terror, but by rendering independent thought structurally optional. Second, it introduces the concept of axiological displacement to describe how AI systems do not merely optimise within existing value frameworks but progressively reconstruct what counts as a value, extending Arendt's analysis of superfluousness into the domain of algorithmic governance. Third, it demonstrates that social media platforms are neurologically preparing the next generation for this delegation, drawing on longitudinal neurodevelopmental research to argue that algorithmic atrophy of prefrontal judgment capacities and the deployment of agentic AI constitute two phases of a single process. The paper concludes that AI governance must move beyond calibration and alignment toward the preservation of institutional spaces for genuine human deliberation, proposing axiological impact assessments as a necessary complement to the existing regulatory framework.