Again, using audio AI to copy a real voice is kinda dumb. Like using image AI to draw real actors, playing characters... instead of just drawing the fucking character.

This tech would let one actor do all the characters. Like how audiobooks work - except you'd get the actor's performance in the character's voice. Stephen Fry doing Snape (behave) may sound like Alan Rickman, but only if Alan Rickman is how Snape's supposed to sound.

The only reason for him to sound exactly like that is if that's what people already think he sounds like. New characters, or newly-adapted characters, can sound however you want. This technology will let anybody play them.

So hire someone good.