Joy Buolamwini

Research

While a graduate student at the MIT Media Lab, Joy Buolamwini found that facial analysis software struggled to detect Buolamwini's face, and went on to study gender and skin-type disparities in commercially sold facial analysis products. The Gender Shades project ran at the Media Lab from January 2017 to August 2020.¹⁴

The 2018 Gender Shades paper with Timnit Gebru evaluated three commercial gender classification systems and introduced a facial analysis benchmark balanced by gender and skin type, after the project found that existing benchmark datasets over-represented lighter-skinned men.⁵⁴

Algorithmic Justice League

Buolamwini founded the Algorithmic Justice League, which uses art and research to show the social implications and harms of AI and to push for more equitable and accountable systems. Buolamwini's research is featured in the documentary Coded Bias, and Buolamwini also makes art about AI's effects on society under the name Poet of Code.¹²⁸

Teaching and writing

Buolamwini's book Unmasking AI: My Mission to Protect What is Human in a World of Machines, published by Random House, blends memoir with an account of research into the harms and biases of AI. Buolamwini's writing has also appeared in publications such as Time, The New York Times, Harvard Business Review and The Atlantic, and Buolamwini publishes a newsletter on Substack.⁸⁶⁷

Work and contributions

• Dr. Joy - Newsletter Show, channel or publication

  Publishes the newsletter, presented on Substack as Dr. Joy - Newsletter by Dr. Joy Buolamwini and linked from the Poet of Code site.⁷²

• Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification Published research

  First author, with Timnit Gebru, of the 2018 FAT* paper that evaluated three commercial gender classification systems and introduced a facial analysis dataset balanced by gender and skin type.⁵⁴

• Unmasking AI: My Mission to Protect What is Human in a World of Machines

  Author of the book, which combines memoir with an account of research into the harms and biases of AI systems and the founding of the Algorithmic Justice League.⁶²

Timeline

1. Jan 2017Gender Shades became an active project at the MIT Media Lab, with Buolamwini on its team.⁴
2. 2018Published the Gender Shades paper with Timnit Gebru at the first Conference on Fairness, Accountability and Transparency.⁵
3. Aug 2020Concluded the Gender Shades project at the MIT Media Lab.⁴

Other documented roles

Roles the sources document that are not confirmed as current.

• Past member (researcher on the Gender Shades project), MIT Media Lab³⁴

Where they work

• Algorithmic Justice League 1 person in the wiki

Related people

• João CarreiraComputer vision · Evaluation and benchmarks
• Richard ZhangComputer vision · Evaluation and benchmarks
• Shusheng YangComputer vision
• Sameer SinghEvaluation and benchmarks
• Jack ClarkEvaluation and benchmarks
• Marina DanilevskyEvaluation and benchmarks
• Niko GrupenEvaluation and benchmarks
• Andrea VedaldiComputer vision
• Alexei A. EfrosComputer vision
• Yann LeCunComputer vision