There's a link to try it out, quite buried in the blog post: https://huggingface.co/spaces/stabilityai/stable-diffusion
Stable Diffusion Public Release
351–360 of 437 posts
Re: Stable Diffusion Public Release
#352Earlier quoted context omitted.
Wait for the 4070, it should be around the perf of a 3080/maybe 4090 for ~500 bucks if rumors hold up. It is coming in a few months. NZXT used to make good pre-builts. Not sure which others have a good rep. DO NOT BUY DELL.
They were saying that before the 3080 was released. It will be cheaper and better than current hardware. Then when it came out you couldn't buy it for a year.
People now go out instead of quarantine and gaming, crypto market crashing, and ETH will go to PoS soon.
Re: Stable Diffusion Public Release
#353Moral question: Such tool will be used to generate lewd imagery involving virtual minors. No way to prevent it upstream by outlawing feeding it real content (whose possession already is illegal). Suffice to add 'childrenize' layer onto adult NN or something. How will the legal system react ? Bundle it into illegal imagery, period ? Maybe it's already the case - I think drawing made public is, not sure. If not, on wha…
Thoughtcrime is not a thing in the US, do not know about other countries https://en.wikipedia.org/wiki/Thoughtcrime Though Think of the children folk have often tried to make it illegal.
It's illegal in other countries, Canada and Australia are especially uptight about it.
Re: Stable Diffusion Public Release
#354Earlier quoted context omitted.
Photoshop requires experience and some talent, this doesn't. If I was some small rebel group in Africa or the Middle East with basically no money or training, I'd use this tool every single day until I was in power, or I'd frame my opposition as using it against the People. Everyone just got their own KGB art department.
Try and do that. Likely those people won't care too much about you, they lean towards authority figures in their community. It is way easier to find those guys and corrupt them. Rather than running some underground news agency changing minds of millions of people. In fact quite often the former is the case + cheaper + less time to execute. "Western" societies will be more resilient to this scenario. So, mostly its go…
>"Western" societies will be more resilient to this scenario
I would think western societies, with our higher consumption of media, would be less resilient. 'Fake news' is already a large issue here.
Re: Stable Diffusion Public Release
#355Played with it for a bit in DreamStudio so I could control more of the settings. So far everything it generates is "high quality", but the AI seems to lack the creativity and breadth of understanding that DALL-E 2 has. OpenAI's model is better at taking wildly differing concepts and figuring out creative ways to glue them together, even if the end result isn't perfect. Stable Diffusion is very resistant to that, and…
It doesn't seem to understand things that DALL-E seems to (usually) understand. For example, it doesn't know that there are 5 fingers on most hands. It cuts objects in half randomly, etc.
I haven't found many of the images it produces to be really usable.
Linked is an example from Dream Studio showing some of these issues: https://i.postimg.cc/qrDKGSVJ/Screen-Shot-2022-08-22-at-7-41...
Re: Stable Diffusion Public Release
#356Played with it for a bit in DreamStudio so I could control more of the settings. So far everything it generates is "high quality", but the AI seems to lack the creativity and breadth of understanding that DALL-E 2 has. OpenAI's model is better at taking wildly differing concepts and figuring out creative ways to glue them together, even if the end result isn't perfect. Stable Diffusion is very resistant to that, and…
I read that you have to give longer prompts to Stable diffusion. This is my bison astronaut: https://i.imgur.com/ohIuG6F.png The prompt was: "A bison as an astronaut, tone mapped, shiny, intricate, cinematic lighting, highly detailed, digital painting, artstation, concept art, smooth, sharp focus, illustration, art by terry moore and greg rutkowski and alphonse mucha"
Not sure how you got that from stable diffusion. This is the kind of thing I get - just a bison.
https://i.postimg.cc/bw1R10gB/A-bison-as-an-astronaut-tone-m...
Re: Stable Diffusion Public Release
#357Earlier quoted context omitted.
According to this post, it needs 6.9 Gb. So the 3070, 3070-Ti, 3080, etc. can all run it. Sadly, my RTX 2060 is below that limit...
Apparently the model decompresses, and it won't fit very well on the 8gb models... I'm willing to give the max settings a spin on my 3070ti, but I'm not very hopeful.
It works on my 2070 8 gb ram.
Re: Stable Diffusion Public Release
#358The most interesting part, to me, of a release like this is the amount of "please don't abuse this technology" pleading. No licence will ever stop people from doing things that the licence says they can't. There will always be someone who digs into the internals and makes a version that does not respect your hopes and dreams. It's going to be bad. As I see it, within a couple years this tech will be so widespread and…
I wonder/hope if a downstream effect of this technological change will be the end of the idea of a "shameful" or "humiliating" image. We all have bodies, we all have sex, and I agree that soon images of ourselves being nude/having sex etc. will proliferate because they'll be generated instantly via Siri shortcut as part of casual banter. In a world where every celebrity is having sex with gorillas, doesn't such an im…
Reminds me of the novel The Light of Other Days by Stephen Baxter: https://en.wikipedia.org/wiki/The_Light_of_Other_Days
I don't generally like his style, but this novel really captivated me. Without spoiling the plot (much), the end-result of everyone being able to spy on anyone anywhere at any time is a kind of societal disinhibition, especially related to sex, nudity, or similar taboos.
Re: Stable Diffusion Public Release
#359DALL-E 2 just got smoked. Anyone with a graphics card isn't going to pay to generate images, or have their prompts blocked because of the overly aggressive anti-abuse filter, or have to put up with the DALL-E 2 "signature" in the corner. It makes me wonder how OpenAI is going to work around this because this makes DALL-E 2 a very uncompetitive proposition. Except, of course, for people without graphics cards, but it'…
Midjourney also completely destroys DALL-e from a price perspective, effectively allowing nearly unlimited generation for approximately $50 a month. Even though DALL-E tends to be better at following prompt details, you're inhibited from being able to explore the space freely because of how prohibitively expensive it can become.
It only costs $30 a month for unlimited Midjourney generation.
Re: Stable Diffusion Public Release
#360The most interesting part, to me, of a release like this is the amount of "please don't abuse this technology" pleading. No licence will ever stop people from doing things that the licence says they can't. There will always be someone who digs into the internals and makes a version that does not respect your hopes and dreams. It's going to be bad. As I see it, within a couple years this tech will be so widespread and…
I wonder/hope if a downstream effect of this technological change will be the end of the idea of a "shameful" or "humiliating" image. We all have bodies, we all have sex, and I agree that soon images of ourselves being nude/having sex etc. will proliferate because they'll be generated instantly via Siri shortcut as part of casual banter. In a world where every celebrity is having sex with gorillas, doesn't such an im…
> doesn't such an image lose its charge?
Yep! Individual images, and individuals in the first place, absolutely loses significance as supplies increase.
> Will Norms and values around sex/body shaming change?
The bar skyrockets! Look at TV stars from before 1995 - they should already appear below-average amateurs to modern eyes.