I’m not knowledge on this subject in general but I do know that on Android you can turn on the Live Caption feature which uses machine learning to process the audio in real-time and puts captions on the screen. It works (to varying degrees) for anything not just YouTube.
Is there some kind of service where the community can currently provide captions for YouTube videos and then it’s just a matter of putting the two together?
EDIT: Apparently YouTube also has a feature to automatically caption videos, again using machine learning. It does have to be enabled on the content creator’s side. I’m assuming you already know about this but would love to hear from you why this may not be appropriate for you.