Mind the hubris: complexity can misfire

Arnald Puy,Andrea Saltelli

2022-06-22

Abstract:Here we briefly reflect on the philosophical foundations that ground the quest towards ever-detailed models and identify four practical dangers derived from this pursuit: explosion of the model's uncertainty space, model black-boxing, computational exhaustion and model attachment. We argue that the growth of a mathematical model should be carefully and continuously pondered lest models become extraneous constructs chasing the Cartesian dream.

History and Overview,General Literature

What problem does this paper attempt to address?

This paper discusses the problems that may arise from pursuing complexity in mathematical modeling. The authors point out that as models become increasingly complex, there are four dangers: explosion of uncertainty space, black boxing of models, exhaustion of computational resources, and model reliance. They believe that the infinite growth of mathematical models should be carefully considered to avoid the models becoming disconnected from the real world in pursuit of Cartesian ideals.

1. Explosion of uncertainty space: Increasing the model parameters leads to an increase, rather than a decrease, in uncertainty because higher-order effects and interaction become more significant in larger hypercubes, increasing the uncertainty of model outputs.

2. Black boxing: The complexity of models can make their behavior difficult to intuitively understand, leading to more code errors and hidden assumptions, resulting in a decrease in the interpretability and reliability of the model.

3. Exhaustion of computational resources: As model details increase, the computational demands rise sharply, and current computing power may not be able to support more complex models, limiting the improvement of model resolution, and posing challenges in data storage.

4. Model reliance: Researchers' dependence on specific models (institutional loyalty) may lead to overlooking applicability and efficiency when choosing models, relying more on habits and convenience.

The paper reminds us that if accepting mathematics as an objective existence is not a concept, then the model may be an abstract metaphor for the real world with inherent uncertainty. Therefore, pursuing more complex models can mislead us in pursuing the "truth" that exists in the minds of modelers. The paper calls for more caution in the modeling process to avoid problems caused by "arrogance," and proposes that relaxing the pursuit of Cartesian ideals may benefit mathematical models.