Co-production, artificial intelligence and replication: the path of routine dynamics

The replication has been a relevant issue in the literature on innovation and technological transfer. From these perspectives, replication is not a simple transfer of templates but also the development and diffusion of social and physical technologies. Research on replication has focused on two major interrelated issues. The first analyzes the directionality of the transfer, from a replicator to a replicant, and vice versa. The second pertains to the organizational space of the replication, whether intra-organizational or inter-organizational. Based on these issues, academic contributions have stressed that in replication processes, there are bidirectional interactions and generative processes between technologies, knowledge, skills, and institutions. The study of these interactions is focused on the micro-level analysis of the replication of organizational routines and evinces the replication dilemma. The replication dilemma refers to the relationship between similarity and singularity in the process of replication. In turn, the increasing incorporation of digital technologies such as algorithms of artificial intelligence requires a debate on the role these artifacts play in the replication of algorithmic systems and organizational routines. Based on all these issues, the objective of this article is to analyze how the process of replication emerges from a co-production of technological solutions (TS) in which artificial intelligence artifacts are interrelated with firms and institutions. Therefore, our perspective goes beyond the studies on transfer that emphasize the linear and unidirectional replicator-replicant relationship in the process of replication. By adopting this approach, we deepen the analysis of co-production processes that involve various types of actors, artifacts, and routines. This article focuses on two in-depth cases of technology transfer processes where the co-production and replication of routines related to AI-based technological solutions give rise to positive feedback, emergent properties, and innovation process. The discussion focuses on theoretical-methodological issues and innovation policy proposals. Three related key points about replication theory and the adoption of TS based on AI in different organizational contexts can be identified from this article. First, the micro-organizational processes of replication considered in this article contribute to a greater understanding of the dilemma between similarity and singularity