I think story and narrative will be the hardest nut to crack. I like to think humans will always be interested and capable of driving at the top-down level.

Image, video, voice, and instrumental will be easy. But it's still going to require a human with very sophisticated understanding of narrative, theme, pacing, structure, etc. to draw these elements together in a pleasing, coherent, and artistic way.

I think voice and performance will still be human-driven by actors for some time, but that a much smaller cast of less-than-supermodel actors and actresses will be able to do everything themselves.

I'll show you current SOTA animation if you email or message me on Discord.