Live data from Hacker News

Stable Diffusion Public Release

stability.ai

231–240 of 437 posts

Re: Stable Diffusion Public Release

#231
post #163

Earlier quoted context omitted.

Curious as to how you arrived at that conclusion.

A quick Google shows an estimate that it was a 100 billion dollar market in 2017. Seems there is a lot of demand.

But what portion of that demand is specifically for artisinally handcrafted authentic sex rather than for apparent sex acts?

Re: Stable Diffusion Public Release

#233

Played 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"

Re: Stable Diffusion Public Release

#234

Earlier quoted context omitted.

Looking forward to a future where all a politician’s quotes are on a blockchain, signed by their private key, and they chose to do so voluntarily out of fear of deep fakes. Will remove all the useless “I didn’t say that”

Do politicians understand blockchains?

Rename it as “verified speech transcripts.” I don’t need to understand video codecs to watch Youtube.

Re: Stable Diffusion Public Release

#235
post #230
post #209

Earlier quoted context omitted.

This is not the same though, it's about being able to deny that that photo of you is real, not that it's taken out of context.

Well what’s a “real” photo? If I take your photo and photoshop (very well) different surroundings. Is it a real photo if you? If I photoshop your face (very well) onto a different body. Is that real? If I feed your photos into a model that can create realistic versions of those photos in different poses or with different facial expressions. Is that real? They all start with something that is very definitely a real ph…

A real photo is one created by photons outside the camera

Re: Stable Diffusion Public Release

#236
post #85

Earlier quoted context omitted.

(OP here) I agree. I am an artist (not by trade but by lifelong obsession) in several different mediums, but also an AI engineer--so I feel a weird mixture of emotions tbh. I'm thrilled and excited but also terrified lol.

It's my fucking job. I've spent my whole fucking life getting good at drawing. I can probably manage to keep finding work but I am really not happy about what this is going to do to that place where a young artist is right at the threshold of being good enough to make it their job. Because once you're spending most of your days doing a thing, you start getting better at it a lot faster . And every place this shit get…

You got to move one step higher and work with ideas and concepts instead of brushes. AI can generate more imagery than was previously possible, so it's going to be about story telling or animation.

Re: Stable Diffusion Public Release

#237
Pretty cool. Although it's interesting that it can't seem to render an image from a precise description that should have something like an objectively correct answer. I tried prompts like "Middle C engraved on a staff with a treble clef" and "An SN74LS173 integrated circuit chip on a breadboard" both of which came back with images that were nowhere close to something I'd call accurate. I don't mean to detract from the impressiveness of this work. But I wanted a sense of how much of a "threat" this tech is to jobs or to skills that we normally think of as being human. Based on what I'm seeing, I'd say it's still got a ways to go before it's going to destroy any jobs. In its current form, it mostly seems like a fun way to generate logos or images where the exact details of the content don't matter.

Re: Stable Diffusion Public Release

#238
post #230
post #209

Earlier quoted context omitted.

This is not the same though, it's about being able to deny that that photo of you is real, not that it's taken out of context.

Well what’s a “real” photo? If I take your photo and photoshop (very well) different surroundings. Is it a real photo if you? If I photoshop your face (very well) onto a different body. Is that real? If I feed your photos into a model that can create realistic versions of those photos in different poses or with different facial expressions. Is that real? They all start with something that is very definitely a real ph…

Simple: a "real" photo is one in which a light field from the real world impinges on a photosensitive media (CCD, film), and directly encodes that information, with some allowance for global light levels, gain, ISO, and speed. Anything else is a modification therein. HDR, multi-exposure compositing, etc, aren't truly "real". They may be 99% real, but aren't 100% real. If you crop it, it's 99.9% real (we have models which can detect cropping and even from which region it originated, obviously it can't reconstruct the missing data).

Yes, by that definition, most photographs already aren't real.

Re: Stable Diffusion Public Release

#239
post #233

Played 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"

Img2img and inpainting is usually a lot easier to get cool results (with a lot of control on your end) with in my view.

Re: Stable Diffusion Public Release

#240
post #33

The 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…

This is great news for those that actually have sex with a gorilla, because now they can claim it's an ai photo. :)

Kidding aside, I think this is actually good. Humans need ephemerality. We are never getting the full version of it back, but with photorealistic ai video and image creation some freedom returns. I think without it a society in which everyone has a camera all the time would mean absolute ossification of social norms. Right now it's very, very new - I mean multiple generations living with the current, or better (eg. recording eye implants) technology.

Post reply on HN