As we get closer to being able to perfectly replicate the sounds of human speech, I wonder if we'll find an "uncanny valley" effect. Check out 0:10-0:13 in the

As we get closer to being able to perfectly replicate the sounds of human speech, I wonder if we'll find an "uncanny valley" effect. Check out 0:10-0:13 in the video (especially the word "now"). I actually prefer this GLaDOS-esque inflection to something perhaps closer to human-producible phonetics. The imperfection in sliding up and down tones is actually very musical, almost like "grace notes".