Does anyone happen to know of any sort of AI or software that can create any kind of NSFW 3D videos of whatever you want?

[TheErotomaniacDreamscaper3685]

Hello. So for quite some time I've been on a little quest to scour across the world wide web for any type of AI that can generate certain NSFW material. But what I'm looking for is an AI NSFW generator that can generate full-motion animated XXX videos with audio, too. Not these so-called "video generators" that create mere slide show videos with generated images that are just BARELY based on your text prompt. Or perhaps some software (preferably free) that can do this for me.

For example: A lady goes to Spooky Island from the 2002 Scooby Doo movie. She gets kidnapped by a demon that breathes a pheromone gas in her face and it makes her horny and wants to get fucked by the demon.

 

[Automate]

I don't know what kind of display adapter you have, but generating even still pictures with AI would require at least an RTX Nvidia card with at least 6 GB of VRAM. That's considered to be the very low end of it. 16 GB is about midrange and some extremely high-end models seem to require 27 GB of VRAM! Video generators available for the public that I have seen only make 4 second small resolution clips (320x320) and very mutated, crazy, inconsistent, random jumpy results, unless you use a real video clip as guidance for the AI, generate frame by frame and then combine the generated frames into a video again. Who knows what kinds of GPU farms you need for making even those. Teaching the AI models for doing that is even more computationally intensive than the generation itself from what I understand.

Additionally, I saw videos on youtube where someone made use of several different types of AI for playing Tomb Raider 1 automatically. Also a voice model of Lara was trained for the purpose and used in-game to have her comment on the environment. The game needed to be paused for several minutes every time a voice comment was generated, since that part of the generation was the most computationally intensive part of it. Here:

https://www.youtube.com/watch?v=0wTf_bbkW2U

You could learn e.g. Blender3d, make some sketch mockup video of the scene, feed those results frame by frame to an image-to-image generator that takes the layout of the mockups as guidance and then combine it back into a video. Of course, audio would still be missing there.

All this is to say that it's probably going to be outside the realm of possibilty to make consistent AI videos for several years still in any public web services, let alone on your own computer.