Specifically I am supposing the superalignment people were generally more concerned about AI safety and ethics than Altman/etc. I don't think this has anything to do with superalignment itself.