Atlas · GenAI 2026
AI Fairness
Algorithmic fairness & bias mitigation
conceptPeak: 2021Explainability & FairnessAI consensus: 2/3
Prerequisites
- hardModel Evaluation
Detecting bias requires measuring disparate impact using metrics (equal opportunity, demographic parity) — metrics literacy is the foundation
Recommended reference
Barocas, Hardt, Narayanan (2023) Fairness and Machine Learning — free at fairmlbook.org; definitive treatment of ML fairness
Notes from AI deep research
Anthropic Opus
Barocas (2023) Fairness and ML = biblia (darmowa). SHAP/LIME. Bias auditing rosnie regulacyjnie
Google Deep Think
Uprzedzenia rasy, płci, poglądów [G#88]
Related skills
- → is subcategory of: AI Ethics(3/3)
- ← is an instance of: Explainable AI(1/3)