Raesetje Sefala, a research fellow at The Distributed AI Research Institute, dives into the transformative potential of AI in post-apartheid South Africa. She discusses innovative data collection methods that can inform government decisions, emphasizing the importance of accurate representation for marginalized communities. Sefala sheds light on how Canada's approach to AI and data transparency can benefit from South Africa's experiences, particularly in enhancing community services and addressing urban governance issues.
The Distributed AI Research Institute emphasizes community involvement in data collection to accurately reflect marginalized neighborhoods' needs and challenges.
DARE aims to influence public policy by providing evidence-based insights that improve access to essential services for marginalized populations.
Deep dives
Using AI to Serve Communities
The Distributed AI Research Institute (DARE) focuses on leveraging technology, specifically AI, to address the needs of marginalized communities. The institute collaborates with these communities to understand how technology impacts them and seeks to identify ways AI can facilitate positive change. By quantifying the harms caused by technology and exploring its potential benefits, DARE aims to empower these communities and ensure their voices are heard in the discourse surrounding AI. The goal is to create methods for data collection that allow governments to better understand and serve the populations that may otherwise be overlooked.
Challenges in Data Collection
The process of gathering accurate data has proven to be more complex than initially anticipated, as many communities are not properly represented in existing governmental datasets. DARE discovered that the government classifications for areas often inaccurately merged marginalized neighborhoods with wealthier suburbs, complicating efforts to identify where services are most needed. This misrepresentation necessitated a grassroots approach, where community members were consulted on accurately labeling and categorizing their neighborhoods. By engaging directly with those affected, DARE aims to create a more accurate mapping of community needs.
The Role of Community Feedback
Community involvement has been crucial in the data collection and labeling process, emphasizing the importance of local knowledge in developing effective AI models. By validating labels through community feedback, DARE can ensure that their data accurately represents the realities faced by these neighborhoods. This collaborative approach not only enhances the quality of the data but also fosters trust between researchers and the communities served. The ultimate aim is to provide actionable insights that can inform government decisions about resource allocation and service delivery.
Transforming Insights into Action
DARE's work extends beyond data collection; they aim to influence public policy and improve access to essential services like healthcare and housing for marginalized populations. By providing evidence-based insights, DARE hopes to guide government planning towards building infrastructure and services that are more aligned with the needs of the communities. The organization also champions the idea that data should be accessible to the public, thereby empowering citizens to hold their leaders accountable. This approach not only serves immediate community needs but also aims for long-term systemic change.
“How do we collect data so that we give it in the hands of the government in an efficient way? We’re in a democratic country. Isn’t this data supposed to inform the government?”
Raesetje Sefala, research fellow at The Distributed AI Research Institute (DAIR), joins to discuss her work analyzing the impacts of South African apartheid using computer vision techniques and satellite imagery. Recorded live at ALL IN in Montréal.
The BetaKit Podcast is presented by Motion: one of Canada's fastest-growing tech companies. Motion is hiring, so check out its careers page. And read The Comeback,a three-part series on how the team’s early experience taught them the importance of solving the right problems, and helped Motion close a $30-million USD financing round.
Get the Snipd podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share & Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered podcast player
Listen to all your favourite podcasts with AI-powered features
Discover highlights
Listen to the best highlights from the podcasts you love and dive into the full episode