I’ve been petty bullish about AI in general but I’m a little uncomfortable with the idea of blocking out the sun. Maybe it would be easier if we just started breeding people to have eight fingers and call it a day?
Brazil gives Meta a few days to remove all chatbots that mimic sexualised conversations by or with children.
2 Likes
evidence above all else that zuck is a fuckin creep through and through. more than any of these egomaniacal billionaire ceo’s that think they’re god. leaks have indicated that even the fuckin product ghouls at meta have been pushing back at him for how aggressively he wants to push shit like this out
1 Like
Absolutely.
And I’m here for Brazil showing that you can actually regulate social media.
The world would be in a better place if governments the world over didn’t surrender responsibility for what happens on social media.
Like it’s very feasible to say that social media companies are completely accountable (read prison) for child porn on their sites, or grooming, or genocide, or teen girls suicide, or sexual blackmail, or neo Nazis.
1 Like
Isso é socialismo!
2 Likes
Isn’t it better for these sick fucks to be chatting with fake kids rather than real kids?
This assumes that chatting with fake kids doesn’t make them more likely to go out and look for the real thing.
1 Like
Does it? I’ve got no idea, but I suspect someone has studied it. Googling down that rabbit hole is not my idea of a good time. Either hypothesis is plausible. It could make them want it more or it could scratch the itch and make them less likely to harm real kids.
I don’t either. But I’d want to know for sure that fake kids doesn’t lead to real kids before I okayed the fake kids.
I assume a lot of people who molest children start out with child porn and then move on to the real thing. And no I won’t be googling that hypothesis.
1 Like
If a non profit medical organization with the right motivation, scientific evidence, and the appropriate safety controls wanted to create a “pedos talk to chatbot kids instead of real kids” program, I would be broadly supportive.
Facebook is 0 for 5 of those criteria.
It’s clear that this feature is a combination of carelessness and a malicious desire to exploit kids and adults more generally. They think that too many restrictions on the bot make it boring.
1 Like
I’m not sure that “sure” isn’t too high a bar. I mean if it probably reduced risk to real kids, would that not be good enough? It’s hard to be “sure” of anything. Also the reality probably wouldn’t likely be that clean. What if it makes 30% more likely to go after real kids and the other 70% less likely? What then?
the ai bots are sexting with literal children, its not pedos sexting with child bots. even more a leaked internal document says this is explicitly allowed under meta ai guidance policy.
if you or i built and deployed something like that, let alone on purpose with the explicit goal of bot sexting kids, we’d go to prison for a very, very long time.
it doesnt seem to me like the people in this conversation know what this is, it’s way worse, lol. meta’s response is “well meta is for 13 and older” like somehow a bot that sexts a 14 year old makes it better vs a 12 year old. like multiple people sat there in a meeting and thought it was a product they wanted to release and then wrote it down. meta’s rampant abuse of children for decades has been out in the open plain as day to see yet very little is done.
Thanks for the clarification, but in that case it sounds like the problem is that the bots are talking to children. Not that they are mimicking children. Even if the bots were pretending to be adults that were sex talking with kids, that would still be a problem.
I think there’s two seperate problems.
Scandal one is the chatbots sending sexy chats to 13 year olds. I believe they’ve changed that.
Scandal two. Brazil is focusing on the ability to build your own AI bot that can do sexy talk while imitating a child
wen guillotines?
Isn’t that encouraging the behavior? I don’t think we should be allowing companies to make money off of facilitating illegal behavior and I’m not sure how letting pedos molest imaginary children isn’t doing just that.
You’re making it sound like this was some big plan that the product owners had in mind, a “feature” that would reduce harm to children or something. This is just an unintended consequence of the rush to release something where there was no thought paid to how such a feature could be abused. If there is any chance this could result in the harm of a child that’s probably reason enough to not just let it free on the internet with no guard rails in place.
2 Likes
The question is whether pedophiles “harming” imaginary children should be a crime at all. As discussed above, it if it keeps them away from real kids (which it may or may not, it seems no one knows and/or wants to make a deep dive to find out), maybe it shouldn’t.
That’s not what I’m doing at all. I’m not opposed to what Brazil is doing given all the facts. I think they are correct. I’m just asking a related question which boils down to: If pedophiles have imaginary kids to occupy themselves, will that be better for real kids? If we get an answer to that, then maybe we can use it in a way that counterintuitively helps kids. And no, I wouldn’t put Meta in charge of that.
Why wouldn’t it be a crime? I understand the slippery slop of “thought crimes” but this isn’t one of those. This isn’t criminalizing what someone thinks. This would be criminalizing behavior that as a society we deem unwanted. I don’t want to make it easier for these folks to satisfy their desires, I want to make it more difficult. At a certain point allowing the behavior is signaling that it isn’t as reprehensible as should be.