Abeba Birhane, a prominent cognitive scientist, currently serves as a Senior Advisor for AI Accountability at the Mozilla Foundation. She also holds the position of Adjunct Assistant Professor at Trinity College Dublin's School of Computer Science and Statistics, where she contributes to the Complex Software Lab. Her research focuses on human behavior, social systems, and the promotion of ethical and responsible AI. Recently, she earned a spot on the UN’s Advisory Board on AI.
Hailing from Ethiopia, Birhane delves into critical race studies, complex adaptive systems, and the nuances of algorithmic bias within her work.machine learning During her collaboration with Vinay Prabhu, Birhane uncovered significant issues regarding the inclusion of offensive and racially biased imagery within extensive image datasets like ImageNet and 80 Million Tiny Images, which are commonly utilized for training AI systems.
Ethiopia is Birhane’s birthplace. She holds a Bachelor of Arts in Philosophy and a Bachelor of Science in Psychology from The Open University. Completing her Ph.D. in 2021, she conducted her research at the University College Dublin's Complex Software Lab after obtaining her Master’s in Cognitive Science in 2015.
2023
In a pivotal 2020 study, Birhane and Vinay Prabhu, the chief machine learning scientist at UnifyID, scrutinized the labeling, categorization, and data collection methods associated with massive image datasets. They found numerous disturbing images alongside sexist and racist labels in widely used datasets such as ImageNet and MIT’s 80 Million Tiny Images, leading to the latter's formal withdrawal by MIT.artificial intelligence Recently, Birhane joined forces with Rediet Abebe, George Obaido, and Sekou Remy to explore the challenges of data sharing in Africa. Their research highlighted how significant power imbalances influence the process, even when data originates from the continent. Their findings were shared at the ACM Conference on Fairness, Accountability, and Transparency.
The Mozilla Foundation has invested a substantial $30 million in Mozilla.ai, a startup dedicated to building a trustworthy, independent, and open-source AI ecosystem, distinct from the influence of major tech companies and academia.
