The Ethical Algorithm, The Science of Socially Aware Algorithm Design by Kearns, Michael -- Read -- Imperial Library of Trantor

Index

Halftitle page Title page Copyright page Dedication page Contents Halftitle page Introduction

Algorithmic Anxiety Sorting through Algorithms Man Versus Machine (Learning) How things can go Wrong Who are we? What this Book is (Not) About A Brief Preview

1. Algorithmic Privacy: From Anonymity to Noise

“Anonymized Data Isn’t” A Bad Solution Breaches and Inferences Smoking may be Harmful to your Privacy A Different(ial) Notion of Privacy How to Conduct Embarrassing Polls Whom do you Trust? Out of the Lab and into the Wild What Differential Privacy does not Promise

2. Algorithmic Fairness: From Parity to Pareto

Bias by Analogy Learning all about you You are your Vector Forbidden Inputs Defining Fairness Accounting for “Merit” Fairness Fighting Accuracy No such thing as a Fair Lunch Fairness Fighting Fairness Preventing “Fairness Gerrymandering” Before and Beyond Algorithms

3. Games People Play (With Algorithms)

The Dating Game When People are the Problem Jump Balls and Bombs The Commuting Game Your Selfish Waze The Maxwell Solution Maxwell’s Equations Cheating on Maxwell Cooperation through Correlation Games Everywhere Shopping with 300 Million Friends Shopping, Visualized A Different kind of Cloud Computing The Echo Chamber Equilibrium Quantifying and Injecting Diversity Medical Matchmaking Algorithmic Mind Games Games Scientists Play (With Data)

4. Lost in the Garden: Led Astray by Data

Past Performance is no Guarantee of Future Returns Power Poses, Priming, and Pinot Noir Scientific Gaming The Sport of Machine Learning Bonferroni and Baidu The Dangers of Adaptivity The Paths Not Taken If You Torture the Data Tending the Garden of the Forking Paths The Garden’s Gatekeeper A Succinct Leaderboard Tending a Private Garden

5. Risky Business: Interpretability, Morality, and the Singularity

Looking for the Light Lighting the Black Box Self-Driving Morality A Singular Threat Clouded Vision What, Me Worry?

Some Concluding Thoughts

One Cause, Many Problems Algorithmic Promise In the Beginning …

Acknowledgments References and Further Reading Index