As sea level rises and flooding events become more frequent, cities are tapping into a range of engineering and planning solutions to reduce flood losses. Beyond traditional ``hard'' infrastructure (e.g., sewage, seawalls), ``soft'' infrastructure that supports and empowers the role of citizens in preparedness has been hailed as a promising avenue for improving adaptation. Despite its promise, there have been very limited attempts to quantify its effectiveness in reducing losses relative to hard infrastructure, particularly in urban environments. This study fills this knowledge gap using generative AI to quantify the effectiveness of a range of hard and soft infrastructure in urban areas. We demonstrate, for the first time, that soft infrastructure can significantly reduce flood losses, with an impact on par with hard infrastructure. Our results open the door to novel ways of confronting adaptation challenges in cities. Soft infrastructure is a viable alternative, especially when budgetary or time constraints make hard infrastructure unattainable. Policymakers should also couple soft and hard infrastructure to maximize resilience.

翻译：暂无翻译