DALL·E and other image generation AI is the future of NSFW game development.

[Nitwitty]

I don't understand why this isn't a thing yet. It's going to be. At some point it's simply going to take over everything. The potential is just too astoundingly great. I'm not an artist but just experimenting with Stable Diffusion for just a little bit has shown how vastly superior AI art is over DAZ it's downright scary. The only thing is that an artist needs to manually fix hands and feet. But more experienced artists needs to chime in on the situation and contribute to the conversation.

By the way, forget DALL-E and Midjourney. It's about Stable Diffusion run locally on your PC. Then again I believe there is something called OpenJourney which can be run locally. The technology keeps evolving.

 

[nokey69]

I don't understand why this isn't a thing yet. It's going to be. At some point it's simply going to take over everything. The potential is just too astoundingly great. I'm not an artist but just experimenting with Stable Diffusion for just a little bit has shown how vastly superior AI art is over DAZ it's downright scary. The only thing is that an artist needs to manually fix hands and feet. But more experienced artists needs to chime in on the situation and contribute to the conversation.

By the way, forget DALL-E and Midjourney. It's about Stable Diffusion run locally on your PC. Then again I believe there is something called OpenJourney which can be run locally. The technology keeps evolving.

There are a few factors I think are at play for this not getting that much adoption YET.

To run Stable Diffusion locally you need a few specific GPUs at the moment, so cost might be one thing.

Then you can always run into various issues that would mayhbe turn off an artist to investigate unlike a developer that might get a turn on investigating what's wrong.

Then you have the difficulty of getting consistency. You can always train a LoHa or Lora to have more consistency across diffusions, but that is also something that an aritst might be turned off by since it can get quite technical to get a good Lora made.

Then you need to use all sorts of tricks either via prompts, size, inpainting, outpainting, etc to create a good scene.

Then you'll get a debate over if it's really art and all sorts of moral conversations of wether to hate the game or love the game that uses such methods for generating content.

All of these are in a constant flux of change as the field has been progressing exponentially over the course of a few months. I'm up for collaborating with anyone that wants to use these methods in creating a game but I think it will take some time for most people to catch up to this tech at the moment.

Iirc OpenJourney is another Stable Diffusion model which tries to create content similar to Midjourney, it's not quite there yet though. For extremely creative stuff Midjourney still has an upper hand.

 

[Nitwitty]

There are a few factors I think are at play for this not getting that much adoption YET.

To run Stable Diffusion locally you need a few specific GPUs at the moment, so cost might be one thing.

Then you can always run into various issues that would mayhbe turn off an artist to investigate unlike a developer that might get a turn on investigating what's wrong.

Then you have the difficulty of getting consistency. You can always train a LoHa or Lora to have more consistency across diffusions, but that is also something that an aritst might be turned off by since it can get quite technical to get a good Lora made.

Then you need to use all sorts of tricks either via prompts, size, inpainting, outpainting, etc to create a good scene.

Then you'll get a debate over if it's really art and all sorts of moral conversations of wether to hate the game or love the game that uses such methods for generating content.

All of these are in a constant flux of change as the field has been progressing exponentially over the course of a few months. I'm up for collaborating with anyone that wants to use these methods in creating a game but I think it will take some time for most people to catch up to this tech at the moment.

Iirc OpenJourney is another Stable Diffusion model which tries to create content similar to Midjourney, it's not quite there yet though. For extremely creative stuff Midjourney still has an upper hand.

Any issues are going to get ironed over time. And by time I mean like tomorrow or next week taking into account how fast things are moving. It's happening. I just don't know when. I'm just surprised attempts haven't been made here yet despite the tools not being quite ready. Or maybe the tools are ready and I'm just not aware of them. I'm still learning.

 

[nokey69]

Any issues are going to get ironed over time. And by time I mean like tomorrow or next week taking into account how fast things are moving. It's happening. I just don't know when. I'm just surprised attempts haven't been made here yet despite the tools not being quite ready. Or maybe the tools are ready and I'm just not aware of them. I'm still learning.

Lots of things are probably in the works, considering how full of NSFW content Civitai is. It is, indeed, just a matter of time.

GPT + Stable Diffusion + some AI voice synthesizer and you have the basis for a game, albeit not necessarily a high quality one up front.

Animation and video is also in the works, as you mentioned, it's just a matter of time.

 

[Nitwitty]

Lots of things are probably in the works, considering how full of NSFW content Civitai is. It is, indeed, just a matter of time.

GPT + Stable Diffusion + some AI voice synthesizer and you have the basis for a game, albeit not necessarily a high quality one up front.

Animation and video is also in the works, as you mentioned, it's just a matter of time.

Yeah I'm waiting for ChatGPT-5 with a voice synthesizer with built in stable diffusion and you just tell it what you want verbally and it just creates it for you without you having to mess with prompt engineering.