I guess it's because if we wanted to hear human voices we would stick with the voice providers/human singers. It doesn't make sense to want a vocal synth sound realistic if it isn't an effort from your own tuning, there's no merit in making it sound realistic when it's realistic by default. And I think V6 reproduced pretty well it's robotic style in their AI vbs, sounds pretty stylish and I really like that (As much as Cevio too that has this kinda out of tune singing style and different from SynthV that's is just so realistic that is boring).
Art is all about style, something can be bad, but would still be loved by the simple fact that it has style. And the fact is that being realistic is a style owned by real singers (And I hope it still that way), we synth enjoyers like Synth songs for other reasons.
Now I have to point out that the take "..., but AI is more welcoming to producers who don't want to deal with all the manual tuning." sounds exactly like an AI bro XD I mean, the manual tuning is the only real job you have inside a vocal synth, if you take the art process out of the experience of doing art you're not really an artist.
Art is all about style, something can be bad, but would still be loved by the simple fact that it has style. And the fact is that being realistic is a style owned by real singers (And I hope it still that way), we synth enjoyers like Synth songs for other reasons.
Now I have to point out that the take "..., but AI is more welcoming to producers who don't want to deal with all the manual tuning." sounds exactly like an AI bro XD I mean, the manual tuning is the only real job you have inside a vocal synth, if you take the art process out of the experience of doing art you're not really an artist.
personally, I've preferred ai updates in most cases (the exceptions being cevio 1.0 vbs) bc i like how each vb has it's own 'singing style'