Dies ist eine Übersichtsseite mit Metadaten zu dieser wissenschaftlichen Arbeit. Der vollständige Artikel ist beim Verlag verfügbar.
Machine Epistemic Singularity (MES) By Jalal Khawaldeh -SUPPLEMENTARY MATERIALS ARCHIVE
0
Zitationen
1
Autoren
2026
Jahr
Abstract
This supplementary materials document accompanies the main study "Machine Epistemic Singularity (MES): A Structural Diagnosis of Scientific Integrity (SDSI) Framework." It provides comprehensive documentation supporting the theoretical, methodological, and empirical dimensions of the SDSI framework across seven integrated parts. Part A contains complete interaction transcripts from three philosophical stress‑tests conducted with generative AI systems (ScholarGPT, DeepSeek‑R1, and Microsoft Copilot) in February 2026, including all prompts, system responses, and analytical observations of five structural tendencies: unattributed recombination, performance without understanding, authoritative posture without accountability, synthetic drift, and recursive reinforcement. Part B presents extended methodological notes detailing system selection criteria, prompt development and validation, parameter standardisation, and coding methodology, including inter‑rater reliability calculations and replication protocols. Part C offers extended theoretical development, including comprehensive literature reviews for the fifteen axes of epistemic vulnerability, formalisation of propagation mechanisms (amplification, accumulation, structural coupling), and conceptual refinement of threshold dynamics (empirical dilution, validation saturation, recursive dependence). Part D provides 24 illustrative scenarios spanning nine professional domains (engineering, medicine, law, education, journalism, marketing, arts, sports, technology), constructed through rigorous multi‑source triangulation and analysed thematically in relation to the SDSI axes. Part E comprises a complete bibliography and annotated references across philosophy of science, metascience, AI technical literature, science and technology studies, journalism, and policy documents. Part F presents a consolidated glossary of all key constructs developed within the SDSI framework, with definitions, primary chapter references, and relations to other constructs. Part G provides a conceptual heuristic diagram illustrating the structural dynamics of the epistemic singularity, integrating the fifteen axes, structural tendencies, and threshold conditions within a visual framework designed to orient inquiry and generate empirical questions. All materials are provided to ensure full transparency, enable critical engagement, and facilitate replication and extension by future researchers. The archive reflects the diagnostic, non‑prescriptive orientation of the main study, offering conceptual resources for epistemic vigilance rather than predictions or prescriptions.
Ähnliche Arbeiten
The global landscape of AI ethics guidelines
2019 · 4.721 Zit.
The Limitations of Deep Learning in Adversarial Settings
2016 · 3.884 Zit.
Trust in Automation: Designing for Appropriate Reliance
2004 · 3.510 Zit.
Fairness through awareness
2012 · 3.302 Zit.
AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations
2018 · 3.200 Zit.