Résumé du livre
Scott E. Page is an American social scientist based at the University of Michigan, where he’s John Seely Brown Distinguished University Professor of Complexity, Social Science, and Management. He’s the author of several other titles, including The Diversity Bonus and The Difference.
The chapter explores the power of models in understanding, predicting, and designing outcomes across various systems. Using the search for Air France flight AF 477 as a starting point, it highlights how models distill complex variables to focus on key elements, enabling practical applications like mapping missile trajectories or species distributions. Models also help simulate future scenarios, allowing for informed decision-making despite inherent inaccuracies. The discussion extends to nonlinear systems, emphasizing the importance of concave and convex functions in capturing real-world complexities, as seen in economic growth patterns or production costs. Linear models, while simpler, are valuable for identifying correlations, though they require caution to avoid conflating correlation with causation. The chapter also contrasts normal distributions, like the bell curve seen in grades, with long-tailed distributions, such as wealth, underscoring the importance of recognizing patterns in different systems. Finally, it advocates for using diverse models to improve accuracy, drawing on Condorcet’s jury theorem to illustrate how combining perspectives enhances reliability, provided sufficient variety is maintained among the models.
Pour lire le reste du livre, téléchargez
Bitely