AIAI alignment, ethical decisions and emotional “aliveness”
A researcher frames their critiques of AI alignment—especially the idea of having “marginal impact”—as work on ethical decision-making, while valuing emotional “aliveness” over pinning down an answer.
TLDR
A researcher connects their critiques of AI alignment to research on ethical decision-making, but says the “aliveness” they praise in a linked essay matters more than settling on an answer. They describe anger as “a kind of emotional aliveness” and welcome that energy. In a follow-up, they accuse most AI capabilities researchers of not even pretending to justify their actions against an external standard. They say many AI safety researchers are not “beyond shame” in the sense the essay describes, and identify those researchers as their intended audience.
Combined views
4.8K
2 Sources, first seen 2h ago
AI alignment, ethical decisions and emotional “aliveness”
A researcher frames their critiques of AI alignment—especially the idea of having “marginal impact”—as work on ethical decision-making, while valuing emotional “aliveness” over pinning down an answer.