The Accuracy-Fairness Dilemma in Machine Learning Machine learning models merely amplify our biases - not eliminate them.