Everitt, 2016 — Safeguarding {AI} Safety: Self-Modification, Utility Preservation, and Corrigibility

DOI: <a href="https://doi.org/10.48550/arXiv.1608.04118" target="_blank" rel="noopener noreferrer">https://doi.org/10.48550/arXiv.1608.04118
URL: <a href="https://arxiv.org/abs/1608.04118" target="_blank" rel="noopener noreferrer">https://arxiv.org/abs/1608.04118

Frames corrigibility, shutdown, or self-modification as a safety problem under capable agency.

Publication links

Frames corrigibility, shutdown, or self-modification as a safety problem under capable agency.