How far can technical formalism take us in truly making an AI system fair? This is the question Tijl De Bie and Maarten Buyl set out to answer in their paper, "Inherent Limitations of AI Fairness", which is featured on the cover of next month's Communications of the ACM (CACM)!
The field of AI fairness aims to measure and mitigate algorithmic discrimination. However, the technical formalism on which this field is built has come under increasing criticism in recent years. We identify eight inherent limitations in the typical, technical approach to AI fairness that inhibit the field's potential to truly address discrimination in practice.
Does this mean we should stop striving for fairness in AI? No, but fairness tools should be transparent about what they can and cannot do. The field may have the potential to make society more fair than ever, but it needs critical thought and outside help to make it happen.
To learn more, you can check out the paper here.
Short on time? You can also watch the video!
