3 ms·

Even though it is possible to generate such images without training on real-life examples, do you deny if an AI were trained on those examples it would be bette

Even though it is possible to generate such images without training on real-life examples, do you deny if an AI were trained on those examples it would be better at this task? Combine that with the fact that it's hard (if not impossible) to tell after the fact whether an AI model has or has not been trained on any one specific image, and it seems like by allowing this you could be unintentionally creating a way to launder real-life CSAM through generative AI.

throwaway14050 3y ago
> Even though it is possible to generate such images without training on real-life examples, do you deny if an AI were trained on those examples it would be better at this task? Nobody is denying that, but that's not the only way to make it better at this task. > seems like by allowing this you could be unintentionally creating a way to launder real-life CSAM through generative AI. There's no way to launder real-life CSAM. If there's a victim, it's a crime, regardless of whether the victim's images went through a ML model or not.
- Ajedi32 3y ago
  > There's no way to launder real-life CSAM. If there's a victim, it's a crime The point is with an AI-generated image you don't necessarily know whether there's a victim or not. You can take an image that wouldn't exist without training data sourced from countless victims, and credibly claim everything is perfectly legal and that there wasn't any victim. That's the danger I'm referring to.
jeroenhd 3y ago
Of course the output would be better, but it's not even remotely required to have such training data. Spending hundreds of thousands training an AI model to generate CSAM sounds like a terrible way to smuggle it across the world. You could feasibly add hidden activation keywords to enable CSAM generation mode without the public noticing, but you could also just put an encrypted file in a model exported as .pickle and it would take years before anyone notices.
- Ajedi32 3y ago
  I said launder, not smuggle. If such models are legal then you can openly sell access to them; even advertise their capabilites. No need to hide anything other than the training data. And because these models would be naturally superior to models not trained on the real thing (as you yourself just admitted) then they would outcompete those models, thus indirectly funding real-life CSAM.

throwaway14050 3y ago

> Even though it is possible to generate such images without training on real-life examples, do you deny if an AI were trained on those examples it would be better at this task? Nobody is denying that, but that's not the only way to make it better at this task. > seems like by allowing this you could be unintentionally creating a way to launder real-life CSAM through generative AI. There's no way to launder real-life CSAM. If there's a victim, it's a crime, regardless of whether the victim's images went through a ML model or not.

Ajedi32 3y ago
> There's no way to launder real-life CSAM. If there's a victim, it's a crime The point is with an AI-generated image you don't necessarily know whether there's a victim or not. You can take an image that wouldn't exist without training data sourced from countless victims, and credibly claim everything is perfectly legal and that there wasn't any victim. That's the danger I'm referring to.

jeroenhd 3y ago

Of course the output would be better, but it's not even remotely required to have such training data. Spending hundreds of thousands training an AI model to generate CSAM sounds like a terrible way to smuggle it across the world. You could feasibly add hidden activation keywords to enable CSAM generation mode without the public noticing, but you could also just put an encrypted file in a model exported as .pickle and it would take years before anyone notices.

Ajedi32 3y ago
I said launder, not smuggle. If such models are legal then you can openly sell access to them; even advertise their capabilites. No need to hide anything other than the training data. And because these models would be naturally superior to models not trained on the real thing (as you yourself just admitted) then they would outcompete those models, thus indirectly funding real-life CSAM.