@hilbertspaess Rather than fixed 'laws', AI safety today relies on training, alignment and continuous supervision — do Asimov’s laws still make sense as a metaphor, or are they now an outdated myth?
@gregisenberg I think there’s one major category missing: AI-native R&D / autonomous scientific discovery. Not just vertical agents doing existing work, but AI systems that discover new materials, drugs, biological systems, engineering designs, etc.