Coordinating Human and Machine Learning for Effective Organizational Learning
通过基于代理的模拟，研究组织如何协调人类学习与机器学习，发现机器学习可减少人类探索性学习需求，人类调整机器学习系统大多有益但可能有害，且依赖机器学习知识有助于动荡环境中的组织学习。
With the rise of machine learning (ML), humans are no longer the only ones capable of learning and contributing to an organization’s stock of knowledge. We study how organizations can coordinate human learning and ML in order to learn effectively as a whole. Based on a series of agent-based simulations, we find that, first, ML can reduce an organization’s demand for human explorative learning that is aimed at uncovering new ideas; second, adjustments to ML systems made by humans are largely beneficial, but this effect can diminish or even become harmful under certain conditions; and third, reliance on knowledge created by ML systems can facilitate organizational learning in turbulent environments, but this requires significant investments in the initial setup of these systems as well as adequately coordinating them with humans. These insights contribute to rethinking organizational learning in the presence of ML and can aid organizations in reallocating scarce resources to facilitate organizational learning in practice.