Envisioning Communities: A Participatory Approach Towards AI for Social Good (2105.01774v2)

Abstract: Research in AI for social good presupposes some definition of social good, but potential definitions have been seldom suggested and never agreed upon. The normative question of what AI for social good research should be "for" is not thoughtfully elaborated, or is frequently addressed with a utilitarian outlook that prioritizes the needs of the majority over those who have been historically marginalized, brushing aside realities of injustice and inequity. We argue that AI for social good ought to be assessed by the communities that the AI system will impact, using as a guide the capabilities approach, a framework to measure the ability of different policies to improve human welfare equity. Furthermore, we lay out how AI research has the potential to catalyze social progress by expanding and equalizing capabilities. We show how the capabilities approach aligns with a participatory approach for the design and implementation of AI for social good research in a framework we introduce called PACT, in which community members affected should be brought in as partners and their input prioritized throughout the project. We conclude by providing an incomplete set of guiding questions for carrying out such participatory AI research in a way that elicits and respects a community's own definition of social good.

Envisioning Communities: A Participatory Approach Towards AI for Social Good

The paper "Envisioning Communities: A Participatory Approach Towards AI for Social Good" critically examines the inadequacies in defining 'social good' within the field of Artificial Intelligence for Social Good (AI4SG). The authors argue against utilitarian frameworks which prioritize majoritarian benefits, often at the expense of historically marginalized communities, and endorse the capabilities approach as a more equitable alternative. This approach emphasizes human welfare equity and posits that an AI project should seek to expand and equalize the capabilities or substantive liberties of the communities it impacts.

Capabilities Approach and AI

Through the lens of the capabilities approach, AI systems should be assessed by their ability to enhance individuals' substantive freedoms to pursue lives they value. The paper discusses how AI can act as a catalyst for social progress by modifying individual capabilities—personal characteristics, access to resources, and social environments. AI projects that prioritize the empowerment of communities can provide tools, enhance existing abilities, and create conducive environments for achieving desired livelihoods.

Participatory Framework: PACT

The authors propose a framework called Participatory Approach to enable Capabilities in communities (PACT), which aligns the capabilities approach with participatory design principles. PACT suggests a shift from solitary design to collaborative processes where community members are integral participants throughout the lifecycle of AI projects. This participatory approach aims to ensure that the research is responsive to the needs of communities by allowing them to define their social good objectives. The paper stresses the importance of involving marginalized stakeholders to collaboratively identify trade-offs and navigate the complexities of project implementations.

Guiding Principles for Participatory Design

The paper provides a series of guiding questions for undertaking participatory AI4SG research:

• Stakeholder Identification: Deciding how community stakeholders will be identified and represented. Particular emphasis is placed on including representatives from historically marginalized groups.
• Sustainability and Compensation: Ensuring stakeholder compensation and sustainability plans for long-term engagement.
• Deliberative Processes: Leveraging consensus-building approaches to reconcile diverse stakeholder preferences.
• Evaluating Impact: Continuous assessment of how the AI system affects stakeholders' capability sets, ensuring community feedback channels to capture evolving impacts, particularly for the most vulnerable groups.

Implications and Future Directions

The PACT framework illustrates an actionable path to integrating ethical guidelines rooted in the capabilities approach with a detailed participatory process. It sets the stage for AI research to be more aligned with equitable social outcomes and aims to shift power toward individuals and communities historically deprived of agency in technological development.

The paper calls for institutional changes and incentives to better support participatory practices as the norm, advocating for reform in publication and evaluation processes within AI research. It underscores the need for interdisciplinary collaboration, long-term stakeholder engagements, and a shift in focus from mere technological advancement to genuinely impactful social change.

Conclusion

By advocating for a participatory framework focused on enhancing community capabilities, the paper contributes to a more robust and morally grounded understanding of AI4SG. It presents a compelling case for AI researchers to prioritize community empowerment not as an ancillary benefit but as a critical criterion for AI projects aimed at social good.