boasidea1

This project is looking into an often overlooked part of the alignment problem: to what do we want to align AI? And how do we define that rigorously so we can objectively measure it? By Boas