It's kind of weird to read an article warning about people who obsessively consume and produce torture content, and find out the article promotes a community that obsessively consumes and produces stalking/doxxing content.
Don't get me wrong, vigilantes are better than serial killers. But kiwifarms doesn't have a reputation for being selective in its targets. I think they're fundamentally caught in the same trap of chasing highs as the people the article describes.
I think automated content moderation filters just aren't set up to detect animal abuse (the article gives an example with Gepini, where the AI doesn't quite seem to detect the depicted abuse). So removing videos needs manual reports, and that doesn't scale.
I'm not sure it belongs on HN, but I'm also reluctant to flag it, because it is well-written and interesting.
Thankfully, the number of online accounts involved appears to be minuscule (assuming the author is credible).
I think "I don't want to know" is a valid reaction to it, for the same reason someone might not want to learn about eg the fine details of ISIS beheadings.
I think the general dynamic is interesting, and we should learn about it, because it's going to be more and more common as time goes on. The internet pushes content towards the extreme.
But the specifics are stomach-wrenching.
While animal cruelty is a component of the Macdonald Triad (along with arson and bedwetting) which supposedly indicates sociopathy (the rigor of the triad as a predictive tool has been seriously questioned) I don't think there's much evidence that the zoosadists consuming this content are would be serial killers. The profile of the average consumer given, of largely middle aged and older women, is also very much not the profile of serial killers, usually men that begin killing in their late 20s-early 30s.
And though the article does make an argument for a link between zoosadism and pedophilia, serial killers are not statistically more likely to be pedophiles than other men. John Wayne Gacy and Jeffrey Dahmer both targeted teenage boys but they're arguably famous because of how abnormal this is/was among serial killers.
> Possible Interpretations:
> Captivity: The monkey is likely being kept in captivity, as evidenced by the harness and the lack of natural surroundings.
> Animal Abuse: The image raises concerns about animal welfare. The monkey's tethering and the apparent lack of suitable living conditions could indicate potential abuse.
> Exotic Pet Trade: The monkey might be part of the illegal exotic pet trade. Many countries have strict regulations against keeping wild animals as pets, especially primates.
The article claims that it means the AI detected the abuse perfectly, but I don't think it did. It points out things that could be signs of abuse (I'm guessing the prompt primed Gemini to look for them), but signs of abuse alone wouldn't be enough to delete a video, especially if the clip was part of, say, a documentary.
The AI missed two things: that the monkey's leash wasn't just a restraint but a form of deliberate torture, and that the purpose of the video is to make a display of the abuse (as opposed to the abuse being an incidental result of bad living conditions). Those are the things you'd really want to flag for deletion, and I think a current tech AI is unlikely to find them without some deliberate prompt engineering.
I do think we're moving towards platforms taking more and more responsibility for this kind of shit, and they should. But it's naive to assume there's some product manager at google somewhere who is aware of the millions of animal abuse videos on the platform and went "oh no we should preserve them, they bring in ad revenue". That shit is niche.