Algorithms, Leadership, Research
Fast and Ethical: Breaking the Speed Limit on Responsible Content Recommendations
Digital media platforms such as Netflix, Facebook, and TikTok are under increasing scrutiny regarding the ethical implications of their personalized content recommendations. To combat bias and avoid skewed content suggestions, sophisticated algorithms can perform additional layers of analysis to ensure that recommendations give space to topics such as racial equity, sexuality, and political persuasion. However, doing this in real time with the conventional algorithmic approach would greatly increase page-load times and create a frustrating user experience. New research affiliated with the Bernstein Center for Leadership and Ethics sets out a new, faster method for applying ethical constraints to produce responsible content recommendations.