A South Korean man has been sentenced to jail for using artificial intelligence to generate exploitative images of children, the first case of its kind in the country as courts around the world encounter the use of new technologies in creating abusive sexual content.
That is not required. Especially in the larger models like a DALLE-3 it can combine concepts even without being directly trained on it. The one they had in the showcase for DALLE-2 was a chair shaped like an avocado. It knows what a chair is and it knows what an avocado is, so it can combine them. So it can know “this is what a naked human looks like” and “this is what a human child looks like” and could combine them without having ever seen CSAM.
See, this is the problem with this entire thread. You guys are, rightfully, upset about what you perceive, but then you take that and use that energy to spread false claims because you think they’re true, or you think the lie will help bolster your side of the argument. Instead it just makes you look disingenuous and paints a bad look for everyone. There’s plenty of accurate points to be made for why this stuff isn’t okay, but you guys are making none of them, very matter-of-factly. Do better. Be better. Make valid arguments, based on fact, or do what people used to do and sit back and let the people who know what they’re talking about make the arguments.
It doesnt, the whole point of generative ai is that it can generalise and mix concepts. There are likely 0 images of astronaughts riding dinosaurs in the training set but it can produce countless pictures like those by generalising the concepts and combining them.
Equally a training set contain pictures of children and adult porn could combine the two.
Exxept for the source csam required to get the model started, of course.
That is not required. Especially in the larger models like a DALLE-3 it can combine concepts even without being directly trained on it. The one they had in the showcase for DALLE-2 was a chair shaped like an avocado. It knows what a chair is and it knows what an avocado is, so it can combine them. So it can know “this is what a naked human looks like” and “this is what a human child looks like” and could combine them without having ever seen CSAM.
Did somebody audit the dataset?
Ya, I thought so…
“I don’t personally know what’s in the data set, so it must include CP” is a breathtakingly pathetic argument. Shame on you.
deleted by creator
So if they audit the inputs and can guarantee no naked kids, you’re fine with this??
See, this is the problem with this entire thread. You guys are, rightfully, upset about what you perceive, but then you take that and use that energy to spread false claims because you think they’re true, or you think the lie will help bolster your side of the argument. Instead it just makes you look disingenuous and paints a bad look for everyone. There’s plenty of accurate points to be made for why this stuff isn’t okay, but you guys are making none of them, very matter-of-factly. Do better. Be better. Make valid arguments, based on fact, or do what people used to do and sit back and let the people who know what they’re talking about make the arguments.
It doesnt, the whole point of generative ai is that it can generalise and mix concepts. There are likely 0 images of astronaughts riding dinosaurs in the training set but it can produce countless pictures like those by generalising the concepts and combining them.
Equally a training set contain pictures of children and adult porn could combine the two.