Work

AI Deployment Across the AI Value Chain: A Dynamic Scenario Model of Infrastructure, Organizational Integration, Human Verification, and End-to-End Net Value GENESIS R94.1b – Research Preprint v0.9

Year: 2026
Type: preprint
Abstract: AI Deployment Across the AI Value Chain: A Dynamic Scenario Model of Infrastructure, Organizational Integration, Human Verification, and End-to-End Net Value GENESIS R94.1b – Research Preprint v0.9 from AI2AIR.Lab Dietmar Fuerste, Oldenburg Germany Abstract The economic impact of artificial intelligence cannot be adequately assessed through model capabilities, benchmark performance, or inference costs alone. Realized value emerges from the interaction of computing infrastructure, organizational integration, task allocation, human expertise, verification capacity, and the consequences of erroneous decisions. This research preprint presents GENESIS R94.1b, a dynamic systems model linking the supply side of the AI value chain—compute capacity, energy constraints, technological efficiency, and pricing—with organizational demand, human–AI task allocation, integration quality, knowledge retention, and risk-sensitive verification. The reference implementation combines 19 dynamic state variables, two explicit feedback mechanisms, and 14 synthetic scenario cases covering the period 2024–2032. Outcomes are evaluated using cumulative end-to-end net value, complemented by separate measures for safety-critical tasks, undetected errors, verification bottlenecks, and control decisions. The simulations reveal several conditional trade-offs: improvements in AI capability do not necessarily translate into proportional economic benefits; aggregate productivity gains may coexist with deteriorating outcomes in safety-critical tasks; additional verification capacity does not automatically improve error detection; and increasing automation can undermine the human expertise required for future oversight and recovery. The study also examines four organizational AI integration architectures—assistance, collaboration, coordination, and control—as conceptual design patterns for human–AI–machine systems. A strategic SWOT assessment for 2027–2031 and contemporary industrial research provide context for potential deployment and adoption pathways. The theoretical contribution lies in integrating AI infrastructure economics, organizational dynamics, task-specific risk, finite human verification capacity, and explicit control mechanisms within a common, auditable simulation framework. The analysis is positioned against existing research on task-based automation, organizational complementarities, AI model routing, and the economics of organizational coordination, including the 2026 working paper Organizing Intelligence. Research status and limitations All scenario results are based on synthetic assumptions and must not be interpreted as empirically validated productivity forecasts, industrial safety guarantees, or operational recommendations. Independent model and manuscript reviews support publication as a provisional research preprint under stated conditions. Two unresolved protection issues in the underlying v0.8c2 implementation prevent a technical architecture or safety freeze. The study contributes a reproducible framework for exploring systemic trade-offs and formulating testable hypotheses for subsequent empirical and technical research. Keywords: Artificial Intelligence; AI Deployment; AI Adoption; AI Value Chain; System Dynamics; Human–AI Collaboration; AI Governance; Human-in-the-Loop; Verification; Organizational Transformation; AI Economics; Industrial AI; End-to-End Value.
Language: English
Open Access status: green
OpenAlex ID: W7235833100
Cites:
Cited by:
Related to: 10