Technically, you can already do that today, as many youtube-videos have now autogenerated subtitles. Someone just need to build a tool utilizing them for this. Though, to be fair, the quality of those subtitles is sometimes a bit questionable. And they are not available in all languages. Especially if you want a different language then what is spoken in the video.

But yes, of course this is another case in which text can shine because of its weakness. Simplicity can be beneficial.