Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

The Cloud Weaving Model for AI development

Darcy Kim, Aida Kalender, Sennay Ghebreab, Giovanni Sileno

TL;DR

The paper introduces the Cloud Weaving Model as an alternative, relational framework for AI development with marginalized communities, drawing on Indigenous and Eastern knowledge to address ethical tensions in mainstream AI. It applies the model to the CommuniCity pilots to reveal patterns of enduring threads, fragile webs, and the tension between sustaining communities and cloud-driven ambitions. A key contribution is a descriptive metaphor that foregrounds community agency, critiques software-engineering centric piloting in social domains, and invites rethinking evaluative frameworks toward more inclusive, ongoing co-creation. Practically, the approach aims to guide more equitable AI development by centering lived experiences, care, and relational accountability, with future work mapping the metaphor to concrete guidelines for ethical design in marginalized contexts.

Abstract

While analysing challenges in pilot projects developing AI with marginalized communities, we found it difficult to express them within commonly used paradigms. We therefore constructed an alternative conceptual framework to ground AI development in the social fabric -- the Cloud Weaving Model -- inspired (amongst others) by indigenous knowledge, motifs from nature, and Eastern traditions. This paper introduces and elaborates on the fundamental elements of the model (clouds, spiders, threads, spiderwebs, and weather) and their interpretation in an AI context. The framework is then applied to comprehend patterns observed in co-creation pilots approaching marginalized communities, highlighting neglected yet relevant dimensions for responsible AI development.

The Cloud Weaving Model for AI development

TL;DR

The paper introduces the Cloud Weaving Model as an alternative, relational framework for AI development with marginalized communities, drawing on Indigenous and Eastern knowledge to address ethical tensions in mainstream AI. It applies the model to the CommuniCity pilots to reveal patterns of enduring threads, fragile webs, and the tension between sustaining communities and cloud-driven ambitions. A key contribution is a descriptive metaphor that foregrounds community agency, critiques software-engineering centric piloting in social domains, and invites rethinking evaluative frameworks toward more inclusive, ongoing co-creation. Practically, the approach aims to guide more equitable AI development by centering lived experiences, care, and relational accountability, with future work mapping the metaphor to concrete guidelines for ethical design in marginalized contexts.

Abstract

While analysing challenges in pilot projects developing AI with marginalized communities, we found it difficult to express them within commonly used paradigms. We therefore constructed an alternative conceptual framework to ground AI development in the social fabric -- the Cloud Weaving Model -- inspired (amongst others) by indigenous knowledge, motifs from nature, and Eastern traditions. This paper introduces and elaborates on the fundamental elements of the model (clouds, spiders, threads, spiderwebs, and weather) and their interpretation in an AI context. The framework is then applied to comprehend patterns observed in co-creation pilots approaching marginalized communities, highlighting neglected yet relevant dimensions for responsible AI development.

Paper Structure

This paper contains 16 sections, 1 figure.

Table of Contents

  1. Introduction
  2. Background
  3. The Cloud Weaving model
  4. Clouds, and their relation to AI
  5. Main elements of the model
  6. Individuality
  7. Dynamism and Flow:
  8. How to weave a web
  9. Differences from physical webs
  10. Application: The CommuniCity project
  11. Pattern 1: Establishing Enduring Threads
  12. Pattern 2: Apps and Platforms Weave Fragile Webs
  13. Pattern 3: Prioritizing Sustenance Over Clouds
  14. Pattern 4: Spiderwebs Breaking and Rebuilding
  15. Pattern 5: No Web Possible
  16. ...and 1 more sections

Figures (1)

  • Figure 1: The five processes expressed in the Cloud Weaving Model: anchoring (by which the spider connects to a cloud), linking (by which the spider connects nearby clouds, although challenged by the weather), bending at the center (by which the spider attempts to find a balance across clouds), spiralling (by which the spider makes the spiderweb more robust, and approaches the center), encountering other spiders (expressing and possibly realizing power relationships).