Sensemaking as a Way to Manage Complexity: Does It Extend to Artificial Agent Organizations?

Designers develop agent-based organizations with the goal to implement systems able to operate in complex environments. The distinctive quality of these systems is that agents are autonomous i.e. they can face complexity through their decision capabilities in order to make it more flexible. This paper proposes that the concept of complexity currently used in artificial intelligence is weak if compared with that elaborated in organizational sciences. As a result, current agent organizations cannot completely face all the range of complex situations. Borrowing some seminal contributions from organizational studies, a broader definition of complexity is presented introducing the notion of ambiguity, and its impact both on reasoning models and organizational design. Moreover, we present some major implications in the design of artificial systems and organizations.