Are there absolutist anti-AI alignment folks? Is aligning AI not to tell teenagers to kill themselves problematic? If I let the devil's advocate part of my brain run wild it can go pretty crazy here.