• 0 Posts
  • 12 Comments
Joined 2 days ago
cake
Cake day: May 6th, 2025

help-circle






  • I … quite literally … showed you it working… flawlessly. Today. Now. The only thing someone needs to do is make the pipeline to do it all automatically. It’s literally already possible, right now. You’re denying what is right in front of you. This isn’t a “in 50 years” thing, this is a … Netflix will probably have auto-dubbed, auto-lip-sync as an option in their library in the next year or less.

    The only thing stopping this from happening would be SAG-AFTRA fighting against the re-dubs because they want their actors to make even more money even though their characters didn’t do any extra work.

    It’s only a matter of time before someone ends up doing it anyways, regardless. This isn’t a futuristic fever dream, it’s already possible NOW.

    Absolutely hilarious that I show you a picture perfect example of its use, and you’re counter-claiming that it’s a pipe dream and not as good as I say…when the proof is literally right in front of you.


  • Free assets are worthless if they don’t match the art style. One of the biggest problems in indie games is assets that don’t match. Free assets only go so far.

    Then there are things like Suno for creating music. etc. You can deny deny deny all you want if it makes you feel better, but you’re already lapping it up without knowing.

    Hell, RealtimeSTT is a great library for doing realtime speech to text, giving people with disabilities the resources needed to participate in media they otherwise couldn’t.

    There’s lots of benefit, not all “AI” are LLMs.

    We’ll be reaching a point soon where foreign film voices are kept, but translated into languages the world-over in the actor/actresses voice, providing unbridled access to media that would have otherwise been unwatched - leading to ever broader cultural reach than before. We’re closer than ever to a universal translator using this technology. Even to the point of matching lipsync to non-native languages.

    https://www.youtube.com/@Thoisoi2 - is a great example of this. His latest videos dub an AI voice over (using HIS VOICE!), and then he also lipsyncs his footage to the voiceover - because he’s not a native english speaker. (he has a THICK russian accent if you watch videos from multiple years ago)

    This video in-particular: https://www.youtube.com/watch?v=LsvLkmNq74s

    This leads to better understanding, clearer speech, more viewer retention, and unless you know he’s doing it - it doesn’t bother you at all. But it’s a HUGE boon for him and his channel.

    Here’s one of his earlier video for comparison: https://www.youtube.com/watch?v=TSzgvP0OW-Q


  • Out of fear of this community retaliating against those examples, I’m afraid not. I do have specific games I could mention, but I’m not willing to point them out to a community who would likely harass said developers.

    However if you’re willing to seek them out, I suggest Godot Developers discord, lots of this going on in Unity/Unreal discord communities as well. AI is being used in a vast number of places to speed up development of artists who would otherwise wade away in tedium to get their jobs done.


  • I’m in a bunch of amateur game dev communities where it’s just teens and 20-somethings with no spare money, and I see it every single hour of every single day.

    The whole “cut of the profits” doesn’t work for artists, because profits of $0 still mean you’re working for free. Most artists won’t oblige that kind of work.

    Lots of these guys are using AI for the general concept, and then having other AI clean up the original work, either by turning it into 3D models, or using mixamo to add animations, etc. I know at least 8 of the indie games I’ve bought rather recently on steam use AI in various places, but with someone touching up the work afterwards to hide any defects.