if one makes AGI via human imitation, and one applies optimization pressure to try to stop it from EVER acting from its analog to the part of human cognition that provides the most robustness…
Posters: davidad. davidad
3 likes recorded by the source corpus.
Topics: Consciousness, AI safety