Whither business model innovation strategy in the age of AI?

Abstract

We argue that business models are transformed into auto-adaptive systems through embedded AI inference and learning activities. The competitive advantage of focal firms may increasingly depend on how effectively their activity systems learn relative to competing business models, with learning serving as a meta value driver. Such learning-based advantages may be difficult to replicate when they are costly to create, path dependent, anchored in distinctive accumulated data, and rely on tightly coupled recursive feedback loops among business model activities. Entrepreneurs and corporate leaders thus become architects and curators of evolving learning systems and innovators of novel business models that are critical to their firm's value creation and sustainable performance. The competitive advantage of focal firms increasingly depends on new AI learning and inference activities, which serve as meta-value drivers in business models that learn faster, more broadly, and more reliably than competing business models. These dynamic learning-based advantages are difficult to replicate because they are costly, path dependent, anchored in specific accumulated data, and rely on tightly coupled recursive feedback loops among business model activities. In this paper we argue that entrepreneurs and corporate leaders need to become architects and curators of evolving, auto-adaptive learning systems and innovators of novel business models that are critical to their firm's value creation and sustainable performance.