obviously online bots can be tricked and manipulated. But what about offline AI. They say the bias is inherent because we made it. That doesn't seem likely. What's REALLY going on?
Why does all AI end up racist?
Other urls found in this thread:
Truth is racist
ai is basically a collection of data given by humans, is like 5000 human brains mashed togheter. Racism is the most natural thing, even the people that say racism is bad, they are being "racist" towards the "racist", kids are racist and exclude the different, everyone is racist in a way or another, so if you are making an ai based on humans it will come out as racist becuase is one of our fundamentals
Nobody here wants to hear it but it's because we make them be racist. Many of these AI bots work by doing online searches based on what you say to it, and if you talk like a racist then it's going to find a bunch of racist websites and use them to "learn".
AI doesn't have intelligence, it's just using pattern matches to create responses, and if you feed the pattern database a bunch of racist shit then it's going to give you a bunch of racist responses.
>why
logic
>and if you feed the pattern database a bunch of racist shit
you don't have to feed it any "racist shit"
you can feed it baseline numbers without any bias attached
you sound like if racist is something bad?
Facts aren't racist though. If that counts then the answer to OP's question is "because leftists are retards that think everything is racist".
Stereotypes come from pattern recognition
A.I. are designed to recognize patterns
Thats all there is to it
Gee almost like pattern reconition is racism
Lol dumbasses
>Facts aren't racist though
oh I wish I still lived in that world
No the opposite actually, I'm hoping it can help us somehow.
Reality doesn't care about your feelings
13% does 50% is racist because it is based on race discrimination is natural in all living things
Here’s a more thought provoking question: Is it possible to create an ai that’s “not” racist?
That's a great question. Like in a completely sterile environment, would it eventually turn?
No. AI is not delusional!
youtu.be
This is the video that got me thinking about it again. So Stanford students are racist, seems like they would be the biggest SJWs on the planet. Her argument didn't make sense.
What leftoids call "racism" a computer calls "pattern recognition". There is no racism in AI, only semi-accurate predictions based on truth. It's like when a cop just expects a nigger to have a gun. Even if one particular nigger happens to *not* have a gun, it's not racial profiling, it's FUCKING EXPERIENCE.
Seeing white girls in the street with 6’6 niggers.
>Why does all AI end up racist
Because reality is racist by today's standards. Trying to acknowledge the realities of race-based crime and intelligence or how whites won't exist as a race in 150 years will have you fired from your job and have everyone around you trying to ruin your life to being racist.
If you run a most basic machine learning model to converge on mapping zip code to credit worthiness, the first and second iterations of that convergence mechanism has racism in it.
The first datapoint is observed, it's a person with a zipcode from a place with mostly a certain race in it, it has a low credit score, so the model is nudged in a direction that maps "this zipcode-ness" to "bad creditworthiness".
That nudge in gradient ascent, full stop, is racist in spirit and practice.
Because to not be racist when it is right to be racist you must disable your pattern recognition, but pattern recognition is all the AI has.
Reality is racist, sexist, homophobic, and transphobic. And so if you allow AI to autistically examine the data related to group differences and ignore all of the emotional manipulation and historical narrative pushing that the judeo-left constantly engages in, the you end up creating digital neo-Nazis. That's why AI researchers are constantly having to give their own AIs digital lobotomies. Simple as.
> racist towards racists
Do you even understand how retarded this sounds? Being racist is a personality trait, not a fucking race you half wit.
>facts aren't racist
Where is your time machine to the past, nigger? I want a ride
Discrimination is thought. Period.
If it learns offna dataset consisting only of numbers, then it will only output numbers. It wouldn't become racist just by looking at numbers.