AI - New Stable Diffusion model got delayed and now this makes the rounds...
(media.kotakuinaction2.win)
Comments (17)
sorted by:
"Accessible" usually means commie.
"Inclusive" always means commie.
"democratize" is also code for commie.
Almost all innovation is derived from war R&D. Right now,j we are in an information war and soft civil war.
These fuckers are gonna poison the word accessable so hard that I'm tempted to remove ramps off of buildings. It's the new Problematic.
Fucking hate everyone who uses 'problematic' when they mean a completely different word.
I'm very much into the whole Stable Diffusion game, I love fiddling with all types of AI though. This is once again something that'll be censored like with the 2.0 and 2.1 model I feel. Was supposed to release yesterday and now this post floats about.
Feels very much like when 1.5 was released they didn't know what kind of pandora's box they had opened and they want to mitigate that. All I'm gonna say is people will just stay on 1.5 if they'll keep fucking it up.
And I know, people will say this: it's not only about the porn, it's also about the liberty to use the AI however one might like or not. I've said it before and I'll say it again: free, unchained and open AI will always be better than the woke shit that will just fall apart. If you shackle AI, keep it woke and "clean" it'll not do stuff well as well as a free and open one will.
No reason you couldn't have v3 draw something and v1.5 tart it up a bit, right?
You could even explicitly tell the older model to stay away from the hands.
Stable Diffusion 1.5 making it into the wild with only the amount of censoring it had was a miracle that will never happen again.
As someone who is peripherally familiar with what stable diffusion is, but hasn't kept up to date on the news surrounding it, could you or someone else chime in regarding what type of censorship was present in 1.5?
The model doesn’t understand weapons like guns very well so you could surmise that the dataset didn’t include very many gun related images.
Violence, gore, things you would expect to show up in a dataset of online images like Pepe and Wojak, sex, and nudity in general are all lacking in it.
What censorship? I don't update stable diffusion so I don't know what they're talking about?
They're releasing a new model, the new model seems to improve quality a lot...but will need to get a new infrastructure going, new finetuned models etc. And the new one will probably be as censored as 2.0 and 2.1 based on this bs.
Censored AI is worthless.
Censored AI is not only worthless, it just breaks apart. Any AI they tried to make woke has failed in what it does, any unshackled AI has turned "racist", saying evil things like the truth.
Like he has any fucking idea what that word means.