AI-based timeline matching

So I had this bizzare idea that I thought I’d put into the universe:

Recently we got some AI love in Articulate (new voices, sound effects, etc…). I use these voices to map rough draft animations and trigger points and stuff on the timeline before recording a human for voiceover. But when I do this I ultimately have to retime all the animations and such to match the slight changes in the voice timing. What if Articulate could do that automatically? What if timeline events could automatically adjust to match the same words/phrases as I change from one voice to another? 
A year ago I wouldn’t have thought of it, but with how good transcription has become in meeting apps I wonder if it’s actually so big a hurdle.