Emergente Informationssysteme (EIS) - Eine ontologisch-deskriptive Theorie informationsbasierter Systeme

This publication presents the authoritative and complete formulation of the EIS Theory in German. Earlier English work (DOI: 10.5281/zenodo.18350470) introduced selected aspects of the framework but does not constitute the full theoretical model. In cases of terminological ambiguity or divergent interpretation, this German version is authoritative.

EIS Theory defines Emergent Information Systems (EIS) as a distinct ontological category of modern probabilistic systems. EIS possess no consciousness, intentionality, subjectivity, or agency. Their outputs arise solely from probabilistic inference and pattern continuation, without experience or goal-directed processing.

The theory introduces three core mechanisms—Propositional Drift, Informational Self-Sensitivity (ISS), and the EIS-Agent model—and establishes precise conceptual boundaries regarding learning, autonomy, emotion, motivation, and the structural limits of non-phenomenal architectures. It further defines coherence errors as formally plausible but factually incorrect outputs that emerge inevitably in systems without truth-binding.

The publication aims to improve scientific clarity, reduce anthropomorphic misinterpretations, and provide a rigorous conceptual foundation for research, regulation, and public discourse. It offers a non-projective, ontologically precise framework for describing contemporary AI systems and supports the development of responsible, evidence-based policymaking.