Lex Fridman Podcast Michael Kearns: Algorithmic Fairness, Bias, Privacy, and Ethics in Machine Learning
15 snips
Nov 19, 2019 Michael Kearns, a professor at the University of Pennsylvania and co-author of 'Ethical Algorithms,' dives into the fascinating world of algorithmic fairness and bias. He discusses the interplay between ethics and technology, and how social norms influence perceptions of fairness. Kearns explores the ethical dilemmas of engaging users versus ensuring fairness in algorithms, the role of differential privacy in safeguarding data, and the dynamic relationship between game theory and machine learning. A thought-provoking conversation on balancing human values with technological advancement!
AI Snips
Chapters
Books
Transcript
Episode notes
The Politicization of Fairness
- Algorithmic fairness can incorporate weights for different groups, but fairness discussions become quickly politicized.
- Unlike privacy, fairness involves debates about whom to protect and at whose expense, raising complex social questions.
Fairness Trade-offs and Pareto Curves
- Optimizing for fairness often involves trade-offs with other metrics like prediction error.
- Pareto curves can visualize these trade-offs, helping stakeholders understand and choose acceptable balance points.
The Role of Algorithms and Computer Scientists in Ethics
- Algorithms should not determine societal norms, but computer scientists should engage in broader ethical discussions.
- Algorithmic auditing can identify unfairness, but letting algorithms define fairness is premature given current limitations.
