Reaction
Alignment engineering vs. misalignment science
A user recommends a LessWrong piece to people working on alignment at AI companies.
TLDR
A user encourages people working on alignment at AI companies to read a LessWrong piece titled “Alignment Engineering” vs. “Misalignment Science.” Its excerpt notes recent discussion about whether a large portion of alignment research is net negative.
Combined views
25.7K
2 Sources, first seen ago