LESSWRONG
LW

2744

"Why Not Just..."

"Why Not Just..."

Aug 08, 2022 by johnswentworth

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

161Deep Learning Systems Are Not Less Interpretable Than Logic/Probability/Etc

3y

56

165Godzilla Strategies

3y

72

106Rant on Problem Factorization for Alignment

3y

53

148Interpretability/Tool-ness/Alignment/Corrigibility are not Composable

3y

13

212How To Go From Interpretability To Alignment: Just Retarget The Search

3y

34

113Oversight Misses 100% of Thoughts The AI Does Not Think

3y

49

82Human Mimicry Mainly Works When We’re Already Close

3y

16

220Worlds Where Iterative Design Fails

3y

30

187Why Not Just... Build Weak AI Tools For AI Alignment Research?

3y

18

155Why Not Just Outsource Alignment Research To An AI?

3y

50

150OpenAI Launches Superalignment Taskforce

2y

40