Summary of Whose Preferences? Differences in Fairness Preferences and Their Impact on the Fairness Of Ai Utilizing Human Feedback, by Emilia Agis Lerner et al.
Whose Preferences? Differences in Fairness Preferences and Their Impact on the Fairness of AI Utilizing Human Feedback
by Emilia Agis Lerner, Florian E. Dorner, Elliott Ash, Naman Goel
First submitted to arxiv on: 9 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computers and Society (cs.CY)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper explores the concept of fairness in content moderation by analyzing how human feedback is used to determine the treatment of comments referencing different sensitive attribute groups. The study uses a novel dataset collected from Prolific and MTurk, which reveals significant gaps in fairness preferences depending on the race, age, political stance, educational level, and LGBTQ+ identity of annotators. Additionally, the research demonstrates that demographics mentioned in text have a strong influence on how users perceive individual fairness in moderation. Furthermore, it is found that differences also exist in downstream classifiers trained to predict human preferences. Finally, an ensemble approach that gives equal weight to classifiers trained on annotations from different demographics outperforms a single classifier for different demographic intersections. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how people decide what’s fair when moderating online comments. It shows that people have different ideas about fairness depending on their own race, age, political views, education level, and whether they’re LGBTQ+. The research also finds that if a comment mentions certain demographics, people are more likely to think it’s unfair. To solve this problem, the study suggests using multiple approaches to predict what people want, each one based on different types of feedback. This way, you get a better idea of what people really think is fair. |
