IBM builds a more diverse million-face data set to help reduce bias in AI – TechCrunch
Encoding biases into machine learning models, and in general into the constructs we refer to as AI, is nearly inescapable — but we can sure do better than we have in past years. IBM is […]