Open Systems Modeling

Takahiro Sasaki

Open Systems Modeling is a methodology that involves understanding the underlying principles through hypothetical thought experiments. It explores fundamental solutions from a broad and long-term perspective. The challenges addressed through Open Systems Modeling include issues such as climate change, biodiversity, food, resources, economic and social stability. Unlike experiments conducted in a laboratory, we do not have complete control over the subject of the problem, since our own inclusion as internal observers means that there are feedback structures inherent in the problem. Moreover, many real-world problems occur only once, and there is no opportunity to reset the situation and start over. In addressing such issues, we view Open Systems Modeling as a methodology that systematically support decision-making using modeling, simulation, and visualization techniques.

The problems we are facing: Issues of Open System.

Humanity is facing various challenges related to planet-wide cyclical failures, such as climate change, depletion of mineral resources, environmental pollution including plastic waste, loss of biodiversity, as well as issues concerning social sustainability such as poverty and inequality. One of the obstacles to resolve these issues is the tendency of individual stakeholders to pursue their own interests in a shortsighted manner. From a global perspective, there has been a broad historical trend of excessive pursuit of economic growth, often resulting in ignoring or externalizing the costs of natural capital behind. Looking at differences within societies, these distortions manifest as regional disparities and intergenerational gaps, attributed to the unfair burden of costs related to human capital and natural resource utilization. Even if some individuals may temporarily reap benefits, in the ultimate sense, given that nature and all societies and economies are interconnected on the shared foundation of the Earth, it will eventually lead to a decline in the overall well-being of humanity, including themselves.

On the other hand, even if well-intentioned measures were implemented, there is no guarantee that the problem will be resolved. In some cases, there may be complexities that could potentially worsen the situation. This is because what we need to address is a systemic problem where the overall behavior arises from relationships between multiple components. In other words, the cause of a problem cannot be attributed to single specific elements constituting the system. The essential cause frequently lies in the relationships between elements and the inadequacy of the overall structure. Moreover, the impact of changes occurring in a part of the system, or changes intentionally induced, often involves time delays before significant effects propagate to other parts, making prediction and control of the system more challenging.

In addressing the challenges we face, a thinking style that goes beyond simple linear approaches is required. Instead, there is a need for a mindset that perceives problems as systems evolving over time.

Simulation not for predicting the future, but for shaping it.

Open Systems Modeling focuses on complex phenomena with strong nonlinearity, where numerous elements are interconnected. It is not centered on deterministic processes dominating simple physical or chemical phenomena. Instead, it particularly addresses societal issues where the influence of highly subjective elements, such as the ambiguity of human intent and behavior or arbitrary factors including political judgments, plays a significant role. Building models for such subjects is often challenging, and there is a tendency to incorporate hypotheses that lack sufficient validation or implement logic based on a certain level of compromise. It is common to question the accuracy of simulations run on models constructed in this manner. Concerns arise regarding the potential introduction of bias by model builders, and the validation of results becomes a critical issue. How can we assert the correctness of simulations conducted on models built in this way? Could the discretion of the model builder impact the results? What methods can be employed to verify the validity of the outcomes? Here, we are questioning the significance of "model building" and attempting a redefinition as follows.

I believe that simulation can be used in two ways. One is as a tool to predict the future using a model that is assumed to be correct, and to foresee and prepare for what may happen. The focus is on eliminating discrepancies between the actual outcomes and the simulation outputs to enhance the accuracy of future predictions, which involves adjusting hypotheses and models. I think this represents a general perspective on simulations in a forward-thinking context.

On the other hand, I also believe another possibility of utilizing simulations in a reverse-thinking manner. In this approach, starting from an envisioned ideal future, the focus is on determining the necessary models to achieve the desired future. To raise the probability of reaching the desired future, we should consider what kind of causal relationships among the system's components are required and within what range of values parameters should be set. This involves exploring what societal measures and institutional designs are necessary and what lifestyles individuals should pursue. Rather than being overly discouraged by the challenges in accurately simulating these factors, it is constructive to accept the potential inaccuracies and view the simulation as a working subject (representing the current state) to explore the conditions required to bring the ideal future closer.

Our perspective is that creating a model doesn't necessarily revolve around making precise future predictions. A model is an expression of the assumptions made by its builder about how things or the world are perceived. Utilizing computer simulations based on these assumptions, models can visualize complex and unpredictable behaviors. We view models as instruments for shedding light on the understanding of phenomena and identifying pathways for addressing challenges and exploring potential avenues for problem solving. Furthermore, considering that models express the perspectives of their builders, we believe they play a crucial role as a common language for exchanging and sharing hypotheses and values among stakeholders who have different knowledge, views and backgrounds. We are actively exploring the application of simulations from this standpoint, and through the practical application to specific societal challenges, we aim to confirm the effectiveness of this approach.