This is what I wanted to say as well. To shift to verbal programming, we'll also have to change to a language designed with that in mind.

Interesting convergence would be VR mixed with TTS processing. So, you use VR and presence/hand tracking to physically interact and words to refine and/or program the visual blocks.