Unmasking AI's Algorithmic Bias

About this episode

This episode explores the insights from Joy Buolamwini's book 'Unmasking AI', focusing on algorithmic bias and its societal implications.

This podcast episode is AI generated, using the tool NotebookLM and summarizes Joy Buloamwini's book "Unmasking AI". Listen in to learn about the groundbreaking insights from Joy Buolamwini's powerful book, Unmasking AI, exploring her personal journey and pivotal research. Key Takeaways from this episode: • AI systems inherently carry algorithmic bias from biased creators and unrepresentative "pale male datasets," leading to disproportionate failures for marginalized groups, especially darker-skinned women. • AI harms extend beyond technical errors, causing profound societal consequences like wrongful arrests and denial of essential services, highlighting the need for a "sociotechnical" understanding of AI. • Algorithmic justice requires shifting power, mandating accountability, ensuring explicit consent for data, and recognizing that not building or deploying certain AI systems is a vital choice. • Driving change in AI requires interdisciplinary collaboration and courageous advocacy, humanizing harms through diverse communication like technical research, art, and personal storytelling, and empowering marginalized voices to influence policy. You can…

People in this episode

Host: nerdnourishment

Topics covered

• AI bias
• algorithmic justice
• sociotechnical understanding
• marginalized groups
• interdisciplinary collaboration

Keywords

• AI
• algorithmic bias
• Joy Buolamwini
• sociotechnical
• marginalized groups
• algorithmic justice
• data ethics

Mentioned in this episode

Books & works: Unmasking AI

More episodes of Data Podcast for Nerds!

• Don't Be Afraid to Break the Code: Revolutionizing IT Audit with Databricks | David Lee · May 28, 2026 · 38 min
• Navigating the Human Side of Artificial Intelligence | Tiankai Feng · May 7, 2026 · 37 min
• From Computer Vision to Children’s Books: Demystifying AI | Kristen Kehrer · March 26, 2026 · 38 min
• Why Clean Data is More Important Than "Sexy" AI Tools | Susan Walsh · March 5, 2026 · 38 min
• 100 Episodes Later... Data Nerd Awards and Unfiltered Data Predictions · February 12, 2026 · 1h 5m
• Highlights from my Third Year as a Podcast Host · January 22, 2026 · 40 min