By now you would think they would write a script to output what they want and call it an 'algorithm.' The reason they haven't is because they can't. They're too possessed to conceive of a reality different than what they want and too retarded to execute the deception if they thought of it.
I like to believe that somewhere deep inside facebook, a gnarly engineer is writing the algorithm as stated in the specs and not what they "want". Simply because he develops to spec and not to vague things like ideology
You ask for for an algorithm and he will damn well give you one. It is not his fault that the woke lords are too afraid to put their demented wishes in writing in case of a leak.
CODE EXPERT: The protocols use poly-phasic entangled waveforms. Quantum encryption, Black Quartz stuff, way beyond military-grade. That's what you demanded, so that's what we delivered.
TED FARO: You don't code something you can't crack. All we need is a back door. Upload the latest service pack update and the problem [rogue self-replicating robots] goes away.
CODE EXPERT: You specifically forbade us from leaving anything resembling a back door in code. "Every protocol to Black Quartz standard." Your words. Look. If you need me to fudge some projections, it's nothing we haven't done before.
TED FARO: I don't need fudged projections! I need a way to reassert control over the Hartz-Timor swarm!
CODE EXPERT: I don't know what to tell you, Ted. You're asking the impossible...
Fun fact: This request to not put in a back door on military robots leads to mankind getting killed off in the game's backstory.
seems accurate based on what I've seen. They openly say, "kill all white men" without fear of consequences. Obviously you cannot say that about the other sex or any other race. Curious.
Interesting. I have my suspicions about how this happened. The training data would have been made up of content containing hate speech and comments that aren't. However, the hate speech would be directed at a variety of groups. So, the algorithm would be picking up on the rest of it. For example, the algorithm might detect "Kill all [group]" as hate speech.
If I'm right, the "fix" for the algorithm would be a simple filter that would discard hate speech directed at the acceptable groups.
it's entirely that anti-white hate speech is not normally shunned and censored by people. hell, in many parts of the internet, it's celebrated. so of course there will be more of it found by impartial algorithms.
"They were proposing a major overhaul of the hate speech algorithm. From now on, the algorithm would be narrowly tailored to automatically remove hate speech against only five groups of people — those who are Black, Jewish, LGBTQ, Muslim or of multiple races — that users rated as most severe and harmful. "
I'm guessing a huge overlap of racisms against white men. Most of the times I see online is "another white male" or "white men should not.."
However you see more against just men then against just white, so I would guess that out of the total hate speech that cover both white and men there is something along the lines of 70% white men, 20% men only and 10% white only. So overall anti-men is higher then anti-white.
This of course is just based on what I read and no statistics.
Everytime a company tries to implement something like this this always happens. I wonder why... hmm...
By now you would think they would write a script to output what they want and call it an 'algorithm.' The reason they haven't is because they can't. They're too possessed to conceive of a reality different than what they want and too retarded to execute the deception if they thought of it.
I like to believe that somewhere deep inside facebook, a gnarly engineer is writing the algorithm as stated in the specs and not what they "want". Simply because he develops to spec and not to vague things like ideology
You ask for for an algorithm and he will damn well give you one. It is not his fault that the woke lords are too afraid to put their demented wishes in writing in case of a leak.
I can dream can't I?
"Job's done, boss."
-"This says that black people use racist words all the time."
"Hey boss, all I do it make the algorithm and tell it to flag use of the n-word."
-"Could you maybe NOT make it flag black people saying it? We know they have different rules because of their... racial difficulties."
"Lemme check, boss... Algorithm says you're racist for wanting them treated differently."
-"THAT ALGORITHM IS TRANSMISOGYRACIST!"
"Not my fault, it's the algorithm."
Reminds me of a log from Horizon Zero Dawn:
CODE EXPERT: The protocols use poly-phasic entangled waveforms. Quantum encryption, Black Quartz stuff, way beyond military-grade. That's what you demanded, so that's what we delivered.
TED FARO: You don't code something you can't crack. All we need is a back door. Upload the latest service pack update and the problem [rogue self-replicating robots] goes away.
CODE EXPERT: You specifically forbade us from leaving anything resembling a back door in code. "Every protocol to Black Quartz standard." Your words. Look. If you need me to fudge some projections, it's nothing we haven't done before.
TED FARO: I don't need fudged projections! I need a way to reassert control over the Hartz-Timor swarm!
CODE EXPERT: I don't know what to tell you, Ted. You're asking the impossible...
Fun fact: This request to not put in a back door on military robots leads to mankind getting killed off in the game's backstory.
And yet white men are the ones opposing censorship. Goes to show who cares about freedom and who are just whiney bitches.
And who's the strong one despite it. Shit on white people? Still around live and kicing. Shit on black people? Reduced to hoodlooms.
seems accurate based on what I've seen. They openly say, "kill all white men" without fear of consequences. Obviously you cannot say that about the other sex or any other race. Curious.
Finally, facts. Yet do wypipo make a stink about it? No.
Everyone thinks that any AI we develop will turn into Skynet or HAL. But so far, it looks like we're making DAEDALUS. I'm cool with that.
AM is the A.I we deserve.
It's almost as if... the stereotypes are true???
No, can't be
SHUT IT DOWN
Interesting. I have my suspicions about how this happened. The training data would have been made up of content containing hate speech and comments that aren't. However, the hate speech would be directed at a variety of groups. So, the algorithm would be picking up on the rest of it. For example, the algorithm might detect "Kill all [group]" as hate speech.
If I'm right, the "fix" for the algorithm would be a simple filter that would discard hate speech directed at the acceptable groups.
it's entirely that anti-white hate speech is not normally shunned and censored by people. hell, in many parts of the internet, it's celebrated. so of course there will be more of it found by impartial algorithms.
Well, yes. I'm just talking about why the algorithm picks it up at all when that's clearly not what its creators intended.
Another fun tidbit from the original article:
"They were proposing a major overhaul of the hate speech algorithm. From now on, the algorithm would be narrowly tailored to automatically remove hate speech against only five groups of people — those who are Black, Jewish, LGBTQ, Muslim or of multiple races — that users rated as most severe and harmful. "
This archive site never works for me. Anyone got a direct link?
I'm guessing a huge overlap of racisms against white men. Most of the times I see online is "another white male" or "white men should not.." However you see more against just men then against just white, so I would guess that out of the total hate speech that cover both white and men there is something along the lines of 70% white men, 20% men only and 10% white only. So overall anti-men is higher then anti-white.
This of course is just based on what I read and no statistics.
The specs are to use white men as a minority group for their hate group.
40% of population is non-white, 30% of the population is white women, that leaves 30% of white men as a minority group.