9 ms·
I haven't followed diffusion image generation development for a while. Where do you find information on what models you can use in the model_ckpt field? Do I ne
by tomalaci 3y ago
I haven't followed diffusion image generation development for a while. Where do you find information on what models you can use in the model_ckpt field? Do I need to import them from somewhere? What are the main differences between them and which are more modern or better?
- nickstinemates 3y agoYou can find them on huggingface, or you can reverse engineer which ckpt you want to use based on an image you've seen generated (like at majin[1] - beware, there's a lot of NSFW/controversial stuff here.) 1: https://majinai.art/ https://majinai.art/
- bavell 3y agoAlso CivitAI but beware the NSFW https://civitai.com/ https://civitai.com/
- CSSer 3y agoSome of this is straight up soft-core child porn. This is fucked up.
- greggsy 3y agoI believe illustrations have been deemed to be abuse material, so I wouldn’t be surprised if LE have started looking into it.
- CSSer 3y agoGeez that’s disturbing. I clicked having no qualms with nudes, artistic or otherwise. I’m not a prude. I’ve seen my fair share of anime girls and AI nudes. Hell, I was raised on the internet before parental settings were a thing, but I didn’t expect that. It’s so gross how it toes a line too.
- dingnuts 3y agothe Fediverse has a big problem with this, too, and I never hear anyone talking seriously about it
- GaggiX 3y agocontext: there are several instances that let you upload illustrations without a restriction on what content the image can contain, it can contain minors, rape, incest. These are instances are between if not the biggest and more active Maastodon instances. The reason why they are so active is because platform like Twitter while having a lot of these stuff too sometimes they will ban your account or at least shadowban you. Many Mastodon instances have banned these ones. In my opinion instances should let the user decide what they find problematic or not and unless it's just spam they shouldn't ban instances.
- unshavedyak 3y agoWhat is a serious talk, though? Ie what can be done? Isn't it a lot like the internet as a whole; report the offenders if you like then move on? To me it's akin to encryption being used for illicit/illegal activities. Any tool that gives power to people can and will be abused by people you'd want nothing to do with. What did you have in mind?
- tekknik 3y agoprobably when someone adds young, childish, and something sexual to the prompt it says no and alerts law enforcement?
- unshavedyak 3y agoSo is there some system that implements text/image classification for automatic law enforcement submission that you're referring to? Also jurisdiction, etcetc. It all seems incredibly complex. Not a reason to "not try", but i suspect we'll struggle to implement even the most basic thing. And even then, take that basic thing and apply it to every software where users can input data. Plus we'd have to convince everyone to do this. Automatic scanning and submission of data is not a well liked topic. Remember how Apple doing basic CSAM scanning was full of panic? Even if a government _forced_ us to do this, jurisdiction alone would be a big question. Some serious questions that need some serious thought, imo. Is being hand-wavy even worth the time?
- GaggiX 3y agoillustrations are not a problem under the law in the United States, but it has to be seen for generated images indistinguishable from reality or almost.
- kleiba 3y agoWho exactly is being abused here? I for one would much rather give pedophiles an opportunity to fulfil their sexual desires through AI-generated pictures than real ones. Of course, we can talk about the training material. Are there actual child porn images in there? I seriously doubt it but who knows? And perhaps a case could be made that AI-generated child porn could be a gateway to invite people who then seek out non-generated material. But I think these are separate discussions to be had.
- gochi 3y agoThey aren't separate discussions, they're directly tied to determining abuse material. Revenge porn is an example of abusive material, despite the subject not being abused in the material usually, they're considered abusive material due to the intent to cause abuse through distribution. So if either case applies, whether it's training based on certain images, or it becomes a gateway, these are discussions to be had directly relating to whether or not it should be classified as abuse material. Additionally, I'm not sure if the recommended help methods by professionals who deal with pedophiles is to let them fulfill their specific fantasies without a care. There are lots of really important discussions to be had, but they're all tied to each other basically. We can't separate them out, nor should we aim to.
- GaggiX 3y agoReading the parent post I think he would believe that revenge porn is abusive material, because there is a person who is getting abused with the distribution of the content, the person didn't consent the distribution. CP is abusive because a child was abused in the creation of the content. The doubt that the author of the parent post has is (from what I understand): who is getting abused with the creation and distribution of a generated image?
- gochi 3y agoThey've brought up 3 ways in which it can be abusive (trained on abusive materials, created for the intent to abuse, or continuing self abuse rather than seeking help), but argued those should be separate discussions vs my thinking which is that these are linked to determining if such material is abusive.
- jay-barronville 3y agoAgreed! I just clicked the link and did a double take. I don't care if it's AI. This is child porn material, and in my opinion, it should be shut down.
- samstave 3y agoYep. There needs to be a REALLY FUCKING STRONG effort to kill all CP AI anything. Full stop. AI should automagically report any attempt at CP.
- colechristensen 3y agoIf you’re running a service you should have automatic filtering, detection, and such. A model by itself though… you might as well ask a pencil to report someone for drawing graffiti. It does not make sense.
- eVeechu7 3y agoModels have been trained on something though. They are not analogous to pencils or brushes.
- GaggiX 3y agoIf you have 4 GB of VRAM, you can finetune a model to whatever you want.
- BHSPitMonkey 3y agoThey are responding specifically to the "AI should automatically report" suggestion. An AI model on its own (without a service built up around it) would not have any mechanisms to use the network, reach out to an FBI hotline, or do anything like this. Protections / limits would have to be added by the operator somehow (or the distributor of models themselves).
- nomel 3y agoI've always assumed this is what will be used to justify the regulation of AI.
- SV_BubbleTime 3y agoThe four horsemen of internet censorship. Money laundering, CSAM, terrorism, drugs.
- jvanderbot 3y agoWhat can someone do about it?
- lopatin 3y agoControversial is one way to put it
- Zuiii 3y agoIt always amuses me when people who think they're the center of the world discover that there are other people with moral takes different than theirs. If no real children were harmed to produce this stuff than it should be treated like any other extreme works of fiction (e.g. violence in video games, graphical descriptions in certain books). Being disgusted is not grounds for banning something lol.
- Zuiii 3y ago> https://majinai.art/ https://majinai.art/ Thank you so much for sharing this. Civitai keeps bugging me to create an account. This doesn't seem to suffer from the same flaw.
- orbital-decay 3y ago>Where do you find information on what models you can use in the model_ckpt field? Do I need to import them from somewhere? You can train (finetune) your own on your reference material.
- smusamashah 3y agoCivitai.com is the current most popular resource for models. Also ckpt format is discourage for security concerns and saftensors is now used instead.