Artificial Intelligence, Data-Driven Learning, and the Decentralized Structure of Platform Ecosystems
指出平台生态系统中，用户数据作为内部资产与去中心化用户基础存在关键区别，导致数据驱动学习可能降低用户价值，且数据质量与用户规模弱相关，对管理者有战略启示。
Gregory, Henfridsson, Kaganer, and Kyriakou (2020) highlight the important role of data and AI as strategic resources that platforms may use to enhance user value. However, their article overlooks a significant conceptual distinction: the installed base of decentralized users who connect with a platform lie outside the boundaries of the platform-owning firm, whereas the accumulated data derived from that installed base exists internal to the boundaries of the firm and under firm control. Accounting for this distinction brings forth two key departures from their theory. First, the decentralized structure of a platform ecosystem makes value capture by the platform an essential consideration when analyzing the implications of data-driven learning for users. Because AI and data allow a platform to increase the share of value the platform owner captures from the users, the value perceived by users can often decline. Second, as an internal asset of the platform firm, data from users and complementors exhibits different dynamics compared with the dynamics that govern the installed base itself. As a result, the quantity and quality of the platform’s stock of data are only loosely coupled with the size of the platform’s installed base. We highlight the strategic implications of this distinction for a manager launching a new multi-sided platform.