Generative AI and Organizational Structure in the Knowledge Economy

The preprint operates within the knowledge-based hierarchy paradigm of organizational structure, focusing on how expertise is distributed and utilized within firms. It examines the impact of a novel technology, Generative AI (GenAI), on this established paradigm.

This preprint exhibits characteristics of both normal science and model drift. It primarily explores GenAI's impact within the existing knowledge-based hierarchy framework, attempting to refine the model by incorporating the unique attributes of GenAI, such as its ability to augment worker capabilities and its inherent risk of hallucinations. This aligns with the problem-solving activities within normal science. However, the preprint also identifies several outcomes that deviate from established theory. For instance, the finding that enhanced GenAI capabilities or reduced hallucination rates can actually narrow the span of control contradicts traditional assumptions about AI's role in organizational flattening. Furthermore, the exploration of GenAI's impact on worker knowledge requirements through "deskilling" presents another divergence, challenging established theories of technological progress. These deviations suggest that the existing knowledge-based hierarchy model may need adjustments to accommodate GenAI's unique characteristics, pointing toward model drift. While not a complete paradigm shift, the preprint hints at potential anomalies that could lead to a future crisis or revolution within the theory of organizational structure. Therefore, classifying the preprint as model drift better captures its contribution to the evolution of the knowledge-based hierarchy paradigm.