I'll allow their use of AI.
Sadly they've got a Ricky Gervais video as the ai subtitling sample, but then again it is his anti Hollywood golden globes speech... which was essentially controlled opposition.
Nevertheless.
I'll allow their use of AI.
Sadly they've got a Ricky Gervais video as the ai subtitling sample, but then again it is his anti Hollywood golden globes speech... which was essentially controlled opposition.
Nevertheless.
The problem with most any technology isn't that it exists, just how it's used. Like I've heard Netflix has started using "AI" to dub content, including adjusting the lip sync, and from basically the moment "AI" became the trend of the day, I was thinking about how it could be used very effectively in dubbing to make content much more accessible across the world — but when this is being done by a for-profit corporation, I'm of course going to be skeptical of it. It isn't exactly being done in the way I would've done it myself.
VLC on the other hand is FOSS developed by a non-profit organization, and it's calling this new feature "AI" because that's a word people recognize. "AI" is above all else a brand: if VLC were just upfront about how there's no real meaningful difference between this new feature and how automated audio transcription and local translation have worked for many years already, at least that I can tell, I guess that wouldn't have looked as "cool" on a booth sign.
Dubbing has been done for like a hundred years, you don't need AI for it.
I was thinking more specifically about using voice cloning to help make dubs in small or endangered languages. Because obviously you don't "need" AI to make a dub — and ideally you really shouldn't — but if one actor alone can be given effectively infinite range, or multiple actors can take turns sharing the same voice, then this could drastically lower the barrier of entry for making a decent-quality dub of something, right? And there might be other use cases, too, but that's the thing that stands out to me.