Library / TED Talks / TED

Back to Library
Joy BuolamwiniSociety, labor, and psychologyBias and real-world harmspublished

How I'm fighting bias in algorithms

Buolamwini demonstrates how dataset and evaluation failures become asymmetric real-world harm, especially when face and identity systems are operationalized in high-stakes settings. The talk connects technical error patterns to structural accountability gaps.

Video

Why this matters

In the sAIfe Hands map, this is a core harms anchor: it grounds alignment discourse in present-day deployment externalities. It helps teams audit not only model performance, but also who bears error costs and who has recourse.

Perspective map

Risk-forwardSocietyHigh confidence
Risk-forwardCaution & harms
MixedBalanced framing
OpportunityUpside & deployment

For band and lens definitions, scoring, and counterbalance: see the Perspective Map Framework in Library methodology.

Risk-forward leaning, primarily in the Society lens. Evidence mode: interview. Confidence: high.

  • - Emphasizes alignment
  • - Emphasizes public harms
  • - Emphasizes harms

References

Related figures

  • Joy Buolamwini
  • Timnit Gebru
  • Margaret Mitchell

Explore further

Counterbalance on this topic

Ranked with the mirror rule in the methodology: picks sit closer to the opposite side of your score on the same axis (lens alignment preferred). Each card plots you and the pick together.

Outbound source

This page is designed to preserve focus inside sAIfe Hands first. Use outbound links only when you want to inspect source material directly.