This is a special post for quick takes by Ratios. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
Mostly the first reason. The "made of atoms that can be used for something else" piece of the standard AI x-risk argument also applies to suffering conscious beings, so an AI would be unlikely to keep them around if the standard AI x-risk argument ends up being true.
There's a wide variance in how "suffering" is perceived, weighted, and (dis)valued, and no known resolution to different intuitions about it.
There's no real agreement on what S-risks even are, and whether they're anything but a tiny subset of other X-risks.
Many people care less about (others) suffering than they do about positive-valence experience (of others). This may or may not be related to the fact that suffering is generally low-status and satisfaction/meaning is high-status.