I use Whisper to generates subtitles on occasion. If you use a tiny model, it is pretty terrible. If you use a large model, it does a pretty good job, even doing phonetic spellings of unusual names. The large model takes some real cpu time to do the analysis, and YouTube isn't gonna burn money for free features.
Yes, a metal machine with no pressure sensors, programmed by some dude in his mom's basement that has never lifted more than 10kg, is sure to be more gentle with luggage.
I throw my kindle into my gym bag and then use it on exercise equipment. It occasionally falls out of the bag. My last one stopped refreshing a section of the screen after 10 years.
I make plenty of devices. I could make an epub viewer, using a low power chip and an e-ink display, but I find my manufacturing tolerances to be less than stringent than Amazon's. I'll pay the extra for a fit and finish that might actually last 13 years such that you have to worry about a major electronics company EOLing it.
The world is going to burn, and our descendants will live in a new stone age, but the shareholders will get a great return on their investments this quarter!