News for hackers. Stuff that matters.
Re:Stable Diffusion 2.0 Release by argella on Wednesday November 23, 2022 @08:59PM (#33727043)
Bummer. AI porn is fun.
Re:Stable Diffusion 2.0 Release by Gigachad on Wednesday November 23, 2022 @09:08PM (#33727101)
The future is probably models trained almost exclusively on porn.
Re:Stable Diffusion 2.0 Release by joshspankit on Wednesday November 23, 2022 @10:35PM (#33727551)
Porn has driven many tech advances. I predict that models trained on specific porn genres will appear as soon as training a good model is doable for under $5000. They’ll get here much quicker if we get video to that mark first.
Re:Stable Diffusion 2.0 Release by snek_case on Wednesday November 23, 2022 @11:08PM (#33727693)
You could probably already get people to pay for a subscription to generate images. Wouldn't be surprised if someone is already working on it.
Re:Stable Diffusion 2.0 Release by rfoo on Thursday November 24, 2022 @05:12AM (#33729603)
The entire NovelAI drama had already demonstrated this.
Re:Stable Diffusion 2.0 Release by moron4hire on Wednesday November 23, 2022 @11:27PM (#33727790)
> Porn has driven many tech advances.

This is an urban myth.

Re:Stable Diffusion 2.0 Release by Godel_unicode on Wednesday November 23, 2022 @11:59PM (#33727944)
Which happens to be true.
Re:Stable Diffusion 2.0 Release by feelandcoffee on Thursday November 24, 2022 @06:42PM (#33737450)
Yes and now. Yeah the VHS VS Beta situation was exaggerated, but you'll be surprised on how much on Netflix, and youtube UI tricks were stolen from innovation made in adult sites.

I'll even say that the high bandwidth push in the public was highly related to that. Even HTML5 video players, adult websites were faster to implement it than big streaming websites that still used flash or similar tech.

Re:Stable Diffusion 2.0 Release by the_shivers on Thursday November 24, 2022 @12:58AM (#33728185)
What tech advances would those be?
Re:Stable Diffusion 2.0 Release by fragmede on Thursday November 24, 2022 @01:29AM (#33728337)
print magazine, cinema, VHS, Internet
Re:Stable Diffusion 2.0 Release by thedorkknight on Thursday November 24, 2022 @10:47AM (#33732631)
How much did porn actually contribute to innovation in any of those though?
Re:Stable Diffusion 2.0 Release by joshspankit on Thursday November 24, 2022 @05:18PM (#33736873)
Let me ask you this in reply:

Have you ever seen a non-porn DVD that had multiple camera angles (a feature defined in the standard DVD spec)?

Re:Stable Diffusion 2.0 Release by guncjy on Thursday November 24, 2022 @05:29AM (#33729714)
And all it took was the ritual degradation and abuse of women.
[dead]
Re:Stable Diffusion 2.0 Release by Gigachad on Thursday November 24, 2022 @01:01AM (#33728197)
I have no idea how it works but I have seen people talking about models trained to draw furry art. And I assume no one spent the millions on AWS to train a full model from scratch.
Re:Stable Diffusion 2.0 Release by jeroenhd on Thursday November 24, 2022 @03:51AM (#33729085)
I believe what they do is take the released version of stable diffusion and then continue training from there with their own image sets. I came across their attempts when looking into how to train the model based on some images of my own; their data set so far reaches between tens of thousands and hundreds of thousands images.

All the difficult parts (poses, backgrounds, art styles) has already been done by the SD researchers, the porn network only needs reference material for the NSFW description/tags/details. This is significantly cheaper.

A similar project, training SD to output images in the style of Arcane, is incredibly successful in replicating the animation style with what seems to be very little actual training data.

I don't think you need to start from scratch at all if you use the SD model as a base, all you need to do is to train it on specific concepts, styles and key words that the original doesn't have.

Re:Stable Diffusion 2.0 Release by version_five on Thursday November 24, 2022 @12:14AM (#33727996)
No. The whole point of these models is that they combine information across domains to be able to create new images. If you trained something just on, say baseball, you could only generate the normal things that happen in baseball. If you wanted to generate a picture of a bear surfing around the bases after hitting a home run, you'd need a model that also had bears and surfing in the training data, and enough other stuff to understand the relationships involved in positioning everything and changing poses.
Re:Stable Diffusion 2.0 Release by vintermann on Thursday November 24, 2022 @05:35AM (#33729761)
Even if porn is what you want, it's not clear that's what you want. It can probably generate better porn if it has a little context about what, say, a bedroom is.

What's more interesting, is that there's evidence (from public posts, I haven't tried these models myself) that models trained on some porn get better at non-porn images too.