LESSWRONG
is fundraising!
"Why Not Just..."
LW
$

"Why Not Just..."

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

150Deep Learning Systems Are Not Less Interpretable Than Logic/Probability/Etc

3y

55

157Godzilla Strategies

3y

71

98Rant on Problem Factorization for Alignment

2y

53

136Interpretability/Tool-ness/Alignment/Corrigibility are not Composable

2y

12

204How To Go From Interpretability To Alignment: Just Retarget The Search

2y

34

103Oversight Misses 100% of Thoughts The AI Does Not Think

2y

49

81Human Mimicry Mainly Works When We’re Already Close

2y

16

207Worlds Where Iterative Design Fails

2y

30

172Why Not Just... Build Weak AI Tools For AI Alignment Research?

2y

18

Review

139Why Not Just Outsource Alignment Research To An AI?

2y

49

Review

149OpenAI Launches Superalignment Taskforce

1y

40