Artificial intelligence (AI) expects to increasingly transform ways in which business is conducted. With change follows a need to question current ways of acting and interacting. Yet, the past becomes the frames through which the future is understood. By drawing on Predictive Brain Theory, which shares the same fundamental underpinnings as the Bayesian brain hypothesis, but uses insights from machine learning and neuroscience, the paper conceptualizes that prospective sense making as a skill to update in-flux network pictures are increasingly required for business managers, which the paper reflects on in the light of AI. The paper provides a novel approach to business managers’ mental capacity in understanding change and in their ability to adapt to structural shifts that require an update on gone-solid assumptions about the business environment, while linking this to AI both as a motor of change, and as challenging the human thought with machine learning.

IMP research is often treated as an empirical perspective describing complexities of repeated business-to-business exchanges and their embeddedness. While building on some common understandings and concepts, this paper asks: How homogeneous is the IMP research? This paper uses cluster analysis to capture the roots and various sub-groups of IMP research as means to depict the question of homogeneity (i.e. a core focus in the research) or heterogeneity (i.e. using references from other fields or specific to sub-fields) of the IMP thought. In this scientific work in progress paper we introduce how we design to use bibliographical methods in order to harvest data from an extensive amount of IMP-related articles written from the 1970’s onwards. In this first attempt to reveal IMP we used overall 294 articles yielded to 10,615 co-citation relationships. A threshold of minimum number of citations of a cited reference was set to five (5) to capture such references that have been cited in multiple publications. We introduce visual mapping of defined subject area clusters and as an example we describe shortly clusters. Perhaps not surprisingly our findings suggest that IMP research is not so homogenous, with at least four clear clusters of IMP-research each utilizing different key references.