Social Choice Theory Offers Framework for Collective AI Control
A new paper proposes a social choice approach to collective control of artificial intelligence, arguing that democratic input should be integrated across the machine learning pipeline—from data collection to alignment. The authors demonstrate that social choice theory provides a mathematically grounded modeling language and axiomatic criteria for evaluating control mechanisms. The work addresses the urgent need for societal oversight of AI systems beyond macro-level governance.
Key facts
- Paper titled 'AI of the People, by the People, for the People: A Social Choice Approach to Collective Control of Artificial Intelligence'
- Proposes collective control of AI grounded in social choice theory
- Argues for incorporating collective input at multiple points in ML development pipeline
- Demonstrates social choice as a modeling language for collective input
- Uses axiomatic methodology to evaluate control mechanisms
- Published on arXiv under Computer Science > Computers and Society
- Submission history available on arXiv
- Current browse context includes references and citations
Entities
Institutions
- arXiv