Loading…
15
I saw a meme about a cartoon dog get flagged for 'violent content'
It was just a picture of that dog from the old cartoon with a funny caption about being tired. I read the platform's note saying it broke rules for 'glorifying violence' which made zero sense. It got me thinking about how the bots scanning this stuff must just be looking for certain images, not context. Has anyone else had a totally harmless image get caught in a filter like that?
3 comments
Log in to join the discussion
Log In3 Comments
loganburns6mo ago
Yeah, the "glorifying violence" tag on a cartoon dog is a perfect example. The filters just match images to a list without any common sense. It's getting worse as they rely more on automated systems.
9
cole_lee96mo ago
Totally agree, it's so broken. The systems just look for shapes and words without any real understanding. A cartoon dog isn't violent, it's silly. They keep cutting corners with cheap AI instead of hiring people who can use actual judgment. It makes the whole platform worse for everyone.
6
roberts.jordan6mo ago
Read an article last week about how these image filters are trained on really basic stuff. They basically just match pixels to a list of banned images, like looking for a specific shape of a gun or a cartoon character. It explains why a silly dog picture gets flagged while real problems slip through. The whole system feels lazy and makes the site worse for normal users.
3