When AI shapes s organizations : The Future of Organizational Design

Digital Twins of Organizations : When AI shapes s organizations

Artificial Intelligence is having or will have an impact on every business in one way or another, in the near future and with an intensity that is difficult to grasp. Everyone believes that their solidly rooted expertise protects them from algorithms which, in their view, will never be able to compete with their experience and know-how. All the more so, since understanding and observing human dynamics, as in my case, is difficult to convert into tokens or bits.

Yes, but for some time now, AI has been penetrating the field of organizational design - my field! And even if there's still progress to be made, the impact is increasingly marked. Here's a roundup of current advances, from ONA to digital twins:

LONA (Organizational Network Analysis): Quantitative Analysis of Organizational Networks relies on software tools to collect data via surveys and analyze networks using metrics such as centrality, link density and betweenness (graph theory). Above all, it is a quantitative method for modeling and analyzing the flow of communications, information, decisions and resources within an organization: ONA identifies the role of people in the organizational network ("Central", "Knowledege brocker", "Peripheral", ...), to offer a vision of the flow of information. One of the main advantages of ONA is its ability to identify areas of friction, silos or congestion in information flows. For example,

Subscribe to DeepL Pro to edit this document. Visit www.DeepL.com/pro for more information.

It can reveal bottlenecks that slow down the flow of decisions or key information. This enables HR managers or executives to act upstream to improve collaboration between teams, reduce tensions and optimize exchanges.

Organizational Network Analysis, From Deloitte

However, this quantitative approach has its limitations. The NAO may, in some cases, reveal sensitive information about team interactions, raising confidentiality concerns. In addition, data collection and interpretation can be biased, which can distort the conclusions drawn. Finally, there is the obvious limitation that what is observed through data does not necessarily reflect the reality of informal exchanges, such as those that take place at the coffee machine, which can nonetheless have a considerable impact on organizational dynamics.

Digital Twins of Organizations: Modeling the Whole Organization

Digital Twins of Things (DTTs) pioneered the virtual reproduction of physical objects to simulate their behavior and prevent possible failures. These digital models are widely used in industry, particularly for testing prototypes before they are manufactured.

Digital Twins of Business Processes (DTBP) go a step further, focusing on organizational processes. These models capture the various stages of business processes. Real-time data analysis helps identify inefficiencies and optimize operations. In logistics, for example, DTBPs can be used to monitor inventories and forecast future requirements with great precision

Digital Twins of Organizations (DTOs) represent a giant step forward in the field of organizational design. Unlike DTT and DTBP, which focus respectively on isolated objects or processes, DTOs aim to model the organization as a whole. They include not only processes and material resources, but also social interactions and human relationships within the organization. This makes it possible to simulate the impact of organizational changes on performance, identify points of friction between departments, or test "what-if" scenarios. DTO would enable the current state of the organization to be reflected in a digital model, and the consequences of interventions to

be predicted, because the idea of DTO is based on three fundamental elements: digital representation of the object, synchronization between reality and the model, and interaction between the model and reality. These digital twins would enable continuous monitoring and analysis of the organization's internal dynamics - a real dream for managers seeking to reduce the uncertainty of transformations.

However, the challenges associated with the creation of complete DTOs are still numerous. Organizations have dynamic sociologies, where the emergent properties of human systems, learning and the diversity of internal realities complicate the faithful modeling of their functioning. Human interactions, conflicts of interest, power plays, feedback or leadership singularities... are difficult to capture and simulate with precision. Uncertainty is omnipresent, and AI's ability to predict or control these aspects is still very limited.

Organizational transformation soon to be entrusted to AIs?

The concept of Digital Twins of Organizations could represent a revolution in the way we design or develop organizations, by entrusting them to algorithms. While DTOs promise a more complete and dynamic vision of organizations, their use still requires significant technological and theoretical evolution. While AI, in particular, is developing in the automation of modeling processes, its ability to understand and reproduce human and social complexity remains limited for the time being... for how much longer ?