Yes, that's correct. It is a slight rebuke, a bit tongue in cheek, but that paper is also old. I grouped them because the groups are, from my perspective, hard

Yes, that's correct. It is a slight rebuke, a bit tongue in cheek, but that paper is also old. I grouped them because the groups are, from my perspective, hard to separate. But there are arguably two camps, "AI safety" and "alignment." Perhaps they should work on that, too, form some camps and argue their cases with something defined instead of masquerading under alignment and AI safety depending on the day. But I could also be totally wrong. Until then, I don't believe either are really operating in reality.