This was from a couple weeks back. We’ve been working hard to clean it up. It’s a constant fight, but we take it seriously.
The CP rings showed up at much higher levels than before with the Brazil bump. We run every image through Thorn’s fingerprinting service. That helps us with attempts to upload the media, but then they advertise telegrams and you have to root them out.
I’ve always been curious where, in your experience, is the demand for this material coming from?
Is it insidious and just totally under the surface everywhere or is it more localized?
Glad to hear you’re taking this stuff seriously, but you may want to be careful talking about this in detail in public without consulting a lawyer. Even if it seems like nothing is wrong with what you’re saying, anything that could be used to imply that you aren’t (or previously weren’t) in compliance with certain moderation standards could be used against BlueSky in court.
[flagged]
"Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith."
"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."
No one wants this material on their platform.
[flagged]
> If you're not even doing basic moderation for the obvious terms in foreign languages
So anyone who starts a website allowing public posting needs to hire moderators fluent in the hundreds of languages that will be used there? There's at least 18 languages that are widely officially used across the world, plus more that are national languages of only one country, and over 7,000 total languages in existence.
This is why we have Section 230. Otherwise, we would only be able to use "approved" languages online to avoid liability, which would obviously be terrible.
Clearly the lesson here is that no-one should try to be a global website. The US version can allow only English and Spanish. The Canada version can allow English and French... However it would be interesting to figure out what they do with e.g. Portuguese written in sound-alike English words (is that possible?)
> anyone who starts a website allowing public posting needs to hire moderators fluent in the hundreds of languages that will be used there?
In the EU, that's how it goes. I think there's a minimum amount of visits/users though
The internet is littered with stories of how hard it is to battle this stuff. Some have even decided it’s not worth it and shut down. This very comments section calls out some efforts a lot more sophisticated than “basic moderation of the obvious terms”. Your cynicism needs some serious calibration.
Their response wasn't great in my opinion, they dodged some of the questions like:
> Why doesn’t Bluesky block known keywords and terms like “child pornography” or “cp”?
> Does Bluesky have different strategies for dealing with text posts discussing child sexual abuse or grooming and images of child exploitation?
> Why doesn’t Bluesky block known keywords and terms like “child pornography” or “cp”?
That's 90s tech and useless today. Not only are long-term accounts never going to post it verbatim (if they want to survive on the platform), it will block both normal communication about the issue and unrelated things (lots of cp in shell scripts). There are way better approaches to detection available.
When anyone tries that anyway, we get the expected results. Like "the event that can't be mentioned" and "unalive" on YouTube.
>lots of cp in shell scripts
Then I guess shell scripts are next.
See, this would have been a good response from them, instead of just not answering the question.
I felt it was obvious and didn't need to be said.
Its a cat and mouse game. Generally, these large companies use information asymmetry as a tool in the war
The first one seems obvious to me, that's like blocking the terms "murder", "rape", "piracy", "theft" ... etc. Normal discussion involves using those terms.