Because morality is definitionally relational. It's impossible to be moral in a circumstance in which there is no one that you could be immoral to.

I think these Venn diagrams overlap, and that you are both correct. Morality and meanness overlap. Having feelings has effects. You can train your feelings and reactions. But also, you're right about morality!

You can train your feelings and your reactions, which is exactly why you *shouldn't* be nice to AI. They are machines whose primary purpose is to trick you into believing they are human beings, even going insofar as to directly lie to you about their interiority.

Treating them as anything other than mindless objects built for the purpose off annoying, deceiving, and ultimately de-humanizing you is to actively participate in your own deception.

Yeah, I agree. Kind of a different point though about the morality and meanness being similar and not identical or whatever. I actually am ambivalent about AI, some are in favor, some are against.

It is a different point, but the difference between morality and meanness is *more* salient in this case, not less.