Hey so I have a legitimate question, we are at the pinnacle of a lot of really groundbreaking new tech, I mean video generation, image to image, image to video, google even just released a new model that allows you to fully explore persistent AI generated worlds where you can like for example paint a wall in a house and leave the house and come back and the wall is painted exactly how you painted it.

So what i'm wondering here is, with all this amazing technology, why can't I easily just for example throw in a manga panel that's black and white, and have it be colored properly exactly the way I want it to be and it looks just as good as you would expect from literally any image generation model, and if I can, why have I not heard about this, why do I not hear about manga artists using this, why is there no tool to streamline this, I feel like this is probably a lot more basic and simple than quite a lot of stuff that exists right?

Do I Just not know? Have I not been spoonfed enough? Everything I look for on this topic, all the tools are fucking shit, or the videos are outdated.

Like surely I would expect to for example, throw in reference data of characters or something or environments that are colored from anime or colored 1 shot panels, and just drop it in and it immediately just does all the work simply without me having to do much at all.

The reason I even ask about this is because I am an avid manga reader, but i'm not gonna pretend that it wouldn't be better in color, and I know for a fact this shouldn't be that hard of a tool to exist.

>ai generated voice overs in the original sounding voice but in any language
that would b cool

>>106159610
I saw someone doing something like this before but only a few lines, they did it very well, I assume it was with elevenlabs, it actually genuinely sounds much better than the western voice actors guild same fucking awful actors that ruin every dub to the pit of hell with horrible localization and horrible acting.

>>106159553 (OP)
Its possible but there are still a lot of limitations, in 5 years I expect these models to be advanced enough that all users will be able to do it even locally on their devices