General-purpose technologies (GPTs), such as AI, have significant economic and societal importance. However, many prevailing theories and models, primarily derived from studying discrete (narrowly applicable) technologies, fall short in addressing GPTs with their distinctive characteristics. A longitudinal narrative study of the IBM Watson case turns attention to profiting from the innovation process rather than profiting from innovation (output), which marks a deviation from earlier theorizing. The study further shows that profiting from innovation process is associated with complementary asset bias and option value of openness, as well as to business and performance risks related to the initial use case selection for developing GPTs that may exhibit notable cost of errors. By addressing special characteristics of GPTs from the point of view of innovators benefiting from innovation, and connecting the general-purpose technology development to temporal aspects of value capture, this study enriches the innovation management literature and informs practical applications.
