alright I got nerd sniped really hard by this So I ended up running some simulations for what happens with varying correlations between alignment signals with true values (rho) and selection…
Posters: theproteindance. theproteindance
2 likes recorded by the source corpus.
Topics: Gender, Software, AI safety
View Post: https://x.com/theproteindance/status/1912273763841999179