The Cloud Weaving Model for AI development

AI development with marginalized communities requires rethinking beyond conventional frameworks. The authors found that standard AI development paradigms were inadequate for understanding challenges in pilot projects with marginalized communities, necessitating a new conceptual model grounded in social relationships rather than purely technical considerations.

The Cloud Weaving Model emphasizes interconnectedness and natural systems. Drawing inspiration from indigenous knowledge, nature, and Eastern traditions, the framework uses metaphors like clouds, spiders, threads, and spiderwebs to represent how AI development is embedded within and dependent on social fabric—suggesting that responsible AI cannot be developed in isolation from community contexts.

Co-creation with marginalized communities reveals overlooked dimensions of responsible AI. The paper's analysis of co-creation pilots identified patterns and considerations for responsible AI development that are typically neglected in mainstream AI frameworks, indicating that engagement with marginalized communities surfaces important ethical and practical insights.

Responsible AI development should account for dynamic, contextual factors. The inclusion of "weather" as a model element suggests that responsible AI frameworks must acknowledge changing conditions, uncertainties, and contextual factors that affect communities—moving beyond static principles toward adaptive approaches.

Indigenous and non-Western knowledge systems offer valuable perspectives for AI ethics. By explicitly incorporating indigenous knowledge and Eastern traditions into the conceptual framework, the paper argues that responsible AI development benefits from epistemological diversity rather than relying solely on Western technical and philosophical traditions.