Not sure how well this holds up. I would very much like to see Scarlet Johansson having very dirty sex. I do not want to watch it if I know an AI created it since it lacks authenticity.
Wanting to see celebrities or coworkers debased is common and perhaps authenticity matters there.
But it doesn't matter much in regular porn usage. I don't see why it would be different for a pedophile. Everything going in your mind still be the same in a pedophile's head, just remove 10 years to the people you imagine.
This is a thing? I don’t know that celebrity and will never see them in real life (probably) (at least not close up), why does anyone care if it’s real or not if it looks real? How does anyone know if any image of a celebrity is real or not at this point? Most of these have been heavily photoshopped/touched up for decades as it is, nude or not; that’s not very authentic either?
I think that a network of trust will be established where "real stuff" can be obtained, but I suspect that the average pedophile is like the average man in that authenticity doesn't really matter indeed.
With the very clear caveat that I'd like to see a world where no one is interested or inclined to abuse children or view CSAM in the first place...
I would like to think you are right. The problem is that we already have a comparable with heroin and methadone. And people still want to take the heroin and not the methadone.
(You can get into somewhat conspiracy theory that methadone is designed to either help someone off heroin quickly or otherwise kill the addict more quickly - either way solving the problem. An addicts of wised up to that and avoid it.)
Following that thought through, won't that make "the real thing" that much easier to hide in a sea of indistinguishable, make it much harder to find and prosecute, ease the restrictions currently on those that want to act out IRL and swap their activities online?
The thrust of the article linked here is that:
the bigger risk is that “AI CSAM is going to bury the actual sexual abuse content,” diverting resources from actual children in need of rescue.
If AI CSAM hides real material in a sea of fake, it means the fake has displaced it and killed its profitability. There will definitely be cases where it will make rescuing victims and persecuting criminals harder, but at the same time it'll rob many of them of the incentive to risk committing the crimes in the first place.
Regardless, I don't know if any of this is preventable. The only strategy that might actually have an effect is for law enforcement to use AI-generated CSAM themselves as a honeypot to periodically "ban wave" consumers and have the hashes of the fakes.
A computer program will enable people to abuse more children? I guess a chatbot might be able to groom more children more quickly than a person. Oh they mean computer generated child porn. I guess if it indistinguishable then it is reportable.
"Think of the children" soon to be used for common sense GPU control.
