Comment by throw_away

I would much rather we use AI to improve subtitling. Better auto speech recognition, per-word millisecond synchronization, positioning based on speaker/screen content-- all these could improve the subtitled experience and would be way less creepy than deepfake voices.