A new paper argues that the process of aligning AI systems with human morality is inherently biased due to developer choices. The research highlights that decisions regarding feature selection, voter sampling, and question framing significantly influence the AI's resulting moral preferences. These choices, often undocumented, can shape outcomes in critical applications like AI kidney allocation, AI agents simulating absent workers, and generative AI depictions of the deceased, suggesting that current aggregation methods alone cannot ensure fair or transparent AI. AI
IMPACT Highlights how developer decisions in AI alignment can introduce bias, impacting fairness and transparency in critical applications.
RANK_REASON Academic paper published on arXiv discussing AI ethics and alignment. [lever_c_demoted from research: ic=1 ai=1.0]
- AI agents simulating absent workers
- AI kidney allocation
- alphaXiv
- CatalyzeX Code Finder for Papers
- DagsHub
- Edyta Paulina Bogucka
- generative AI depictions of the deceased
- Gotit.pub
- Hugging Face
- ScienceCast
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
