Questions and Discussion
Talkat Wrong at the Root: Racial Bias and the Tension Between Numbers and Words in Non-Internet Data · 2019-06-05..06-07·
summary
Argues that building a risk model to risk stratify families in one county produces targeted service bins but also correlates with higher inadvertent child removals and offers no pan-county remedy; cites tool details like Boston police gang metrics (features include classmates, colors, schools) and the Chinese social credit model as incentive-driven examples, plus Canvas faculty ranking and a bone-density scanner that misclassifies Black patients due to sparse data; claims legal standards and transparency tools alone do not fix governance or accountability.
AI-written summary · about
chapters (4)
related talks
01h 04mData Hemorrhage, Inequality, and You...
40mPersonal Threat Models and, Speculatively, Privacy Governance
39mEthical ML: Who's Afraid of the Black Box Models?
31mEthics of algorithms - from theoretical discussion to practical application
28mMeasuring Model Fairness
34mEthical Data Collection for Regular Developers
23mDeveloping an intersectional framework to analyze biases in artificial intelligence and deep neural
54mEthical Machine Learning