Back to search

National and Kapodistrian University of Athens

Artificial intelligence, between the global and the national: beyond fictional metrics

Abstract

dc:description

Artificial Intelligence (AI) is increasingly embedded in the fabric through which societies produce knowledge, govern institutions, and pursue economic and social objectives. In this dissertation I examine what it means to measure AI under these conditions of integration, and how the integration of AI into scientific and technological practice challenges the ontological and epistemic assumptions of conventional science and technology (S&T) indicator frameworks. I argue that “measuring AI” is not a purely technical practice in indicator selection, but a socio-technical and governance-oriented endeavor about what becomes visible, actionable and contestable in policy and institutional decision-making - and what remains structurally obscured. Chapter 2 and Chapter 3 situate this argument historically by tracing how S&T indicators emerged as instruments for governing complex knowledge systems, became institutionalized through international organizations (notably the OECD, UNESCO, and the European Union), and were subsequently problematized by STS scholarship and responsible-metrics debates that emphasize the performativity of indicators and their role in governance by metrics. In Chapter 4, I first delineate how AI is framed in global policy discourse i.e. in established international bodies, specialized expert groups and national strategies, and show how these international frameworks and national narratives stabilize certain understandings of AI and in turn shape what is considered as measurable. I then examine the ecosystem of global AI indicator frameworks, describing their conceptual and methodological underpinnings. In Chapter 5, I compare these national narratives/topics with the conceptual (in terms of what) and methodological (in terms of how) frameworks of leading global AI indicator frameworks. For conducting this, I use a mixed-method approach that combines quantitative topic modelling with qualitative interpretation to identify dominant topics and policy priorities. This analysis reveals a structural misalignment. While national AI strategies increasingly emphasize ethical commitments, societal impacts, and nation-specific challenges (i.e., health, education, agriculture), global indicator frameworks continue to privilege capacity-oriented proxies that are easy to standardize such as investment levels, talent pipelines, and volumes of research and innovation output, thereby co-producing a narrower policy understanding of progress in the age of AI. In Chapter 6, I examine how AI is reshaping the processes and practices of scientific knowledge production. Through a synthetic literature review, I analyze how the integration of AI into research, publishing, and evaluation challenges established practices - ultimately reshaping foundational notions of authorship, labor, data and compute dependencies, and validation. The review results indicate that AI disrupts not only the operational adequacy of conventional S&T indicators (publications, citations, patents, and R&D expenditure), but also their epistemic authority, weakening the assumed link between these proxies and genuine scientific contribution. In light of such shifts, I review emerging proposals for new, AI-aware metrics that not only acknowledge the limitations of conventional S&T indicators but also treat ethical grounding and societal responsiveness as integral dimensions of what constitutes progress in the age of AI. In Chapter 7, I explore the role of science in addressing what is commonly referred to in literature as existential risks associated with artificial intelligence. I then argue that prevailing measurement paradigms should evolve to evaluate AI-enabled research in terms of its contribution to broader societal priorities, including the Sustainable Development Goals (SDGs). In Chapter 8, considering AI’s systemic risk profile and the growing debate around existential risk horizons, the dissertation advances an AI-aware measurement paradigm oriented toward public value, responsibility, and anticipatory governance. I translate this argument into policy-relevant recommendations for Greece, proposing a national AI Observatory that moves beyond the metrics paradigm to integrate three interrelated components: (i) Sustainable Development Goals (SDG)-oriented measuring and monitoring of AI-enabled research pathways, (ii) risk-oriented measuring and monitoring via scenarios and institutional readiness signals, and (iii) incident-based measuring and monitoring of local AI use and misuse, feeding back into preparedness and oversight. Crucially, I argue that this architecture can only function as anticipatory regulation if it is designed as a participatory governance infrastructure that translates contested values into indicators and policy triggers through public engagement, thereby making early warnings democratically legitimate and practically actionable. Through these contributions, this dissertation aspires to provide a pathway for reorienting science and technology measurement in the age of AI. It argues that moving beyond fictional or narrow capacity metrics requires treating measurement as a reflexive socio-technical and governance-oriented endeavor - one that is ethically embedded, societally responsive, and capable of capturing the complex transformations of knowledge production. By diagnosing the politics of conventional indicators, mapping their misalignment with policy narratives, systematizing AI’s disruptive effects on research processes, and proposing a concrete observatory mechanism for Greece, I develop a measuring and monitoring framework that opens up the black box of metrics, holding them accountable not only to technological change but to the full spectrum of societal, epistemic, and normative complexities that AI brings into focus.

Degree

thesis:*
Grantor dc:publisher
National and Kapodistrian University of Athens
Year dc:date
2026

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Sioumalas-Christodoulou, Konstantinos
  • Σιουμάλας-Χριστοδούλου, Κωνσταντίνος

Subjects

dc:subject × 21

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/61542

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Sioumalas-Christodoulou, Konstantinos; Σιουμάλας-Χριστοδούλου, Κωνσταντίνος. Artificial intelligence, between the global and the national: beyond fictional metrics. National and Kapodistrian University of Athens, 2026. http://hdl.handle.net/10442/hedi/61542