Praise Jesus it returned… but apparently it was “political.” I guess they aren’t a fan of equality for you and me:
And the Declaration of Independence well that’s hate speech:
2 Likes
Seriously though, is QA that expensive?
whats facebook? ;p
1 Like
My wife and I talked about this the other day. She pointed out that the Declaration of Independence does contain phrases that would indeed be considered hate speech out of context:
He has excited domestic insurrections amongst us, and has endeavoured to bring on the inhabitants of our frontiers, the merciless Indian Savages whose known rule of warfare, is an undistinguished destruction of all ages, sexes and conditions.
There’s more, but just to point it out. Can’t really blame an algorithm when it’s doing what it’s supposed to do? Also, Facebook’s algorithm has been doing this for years, it’s only just coming into the limelight because of all their major screw-ups recently.
Disclaimer: I haven’t been on Facebook for a year and a half because I bothered to read the privacy agreement and saw all the invasion of privacy blatantly outlined.
1 Like
Considering FB returned the post its pretty obvious it is not doing what it is supposed to be doing.
We also talked about how it would be difficult to circumvent. If you whitelisted phrases straight from historical documents, you would end up with people piecing together their hate speech like a ransom note from a newspaper.
You obviously think very highly of programmers to think anyone could knowingly prevent things like this from ever occurring. Perfect code is nonexistent, and algorithms are difficult to guide in many cases.
EDIT: sounded brash before edit.
Never mentioned “perfect code” but thank you for putting words in my mouth. How about NOT relying upon code… but actually training people to review B E F O R E you pull content, not A F T E R.
https://zephoria.com/top-15-valuable-facebook-statistics/
TLDR: Over a billion daily users, 300 million photos uploaded per day, 4.75 Billion pieces of content shared per day.
Good luck doing that by hand.
This is why you do it by algorithm, but have the final removal checked by human, first.
If you can’t do that then either:
- don’t attempt to censor at all, unless reported first
OR - don’t permit photo upload.
I like the “no attention unless flagged”-attitude.
1 Like
What is sad about these algorithms is how they will be relied upon and not questioned. Right now, tech giants use them to try to keep their platforms sanitary. But soon there will be a time when people will actually be arrested and convicted by algorithms. There will probably be a person in the loop to prevent miscarriages of justice, like a judge, but they will get so used to the algorithm being right 97% of the time, human oversight will be nothing more than lip-service.
The US government already uses metadata and algorithms to decide who to drone strike, and because it’s not happening here yet, nobody cares. I’m not sure what human oversight those programs offer, but it is probably already lip-service as well.
I have never used facebook, or twitter, or instagram, but I do know they have a profile on me and everyone else. Stories like these only enforce my belief in the pitfalls of our future reliance on technology. Human beings themselves already have a difficult time deciding things, and people on the whole seem to be dumber than a bag of hammers. Putting computer programs in charge of things will always result in these kinds of errors. Even putting real people in these decison-making roles results in errors.
When companies like facebook sell police versions of these algorithms, we are all doomed.
last I knew, facebook, google, twitter, snapchat, instragram, apps like that, are lifelog. that thing no one heard of when beating the drums to war on so called, terror.