LESSWRONG"Why Not Just..."
LW

"Why Not Just..."

Aug 08, 2022 by johnswentworth

A compendium of rants about alignment proposals, of varying charitability.

116Deep Learning Systems Are Not Less Interpretable Than Logic/Probability/EtcΩ

6mo

52

149Godzilla StrategiesΩ

6mo

63

73Rant on Problem Factorization for AlignmentΩ

4mo

48

111Interpretability/Tool-ness/Alignment/Corrigibility are not ComposableΩ

4mo

8

138How To Go From Interpretability To Alignment: Just Retarget The SearchΩ

4mo

30

87Oversight Misses 100% of Thoughts The AI Does Not ThinkΩ

4mo

49

68Human Mimicry Mainly Works When We’re Already CloseΩ

4mo

16

138Worlds Where Iterative Design FailsΩ

3mo

16