Why everytime scientists make an AI they have to shut it down because is very racist againts blacks?

Why everytime scientists make an AI they have to shut it down because is very racist againts blacks?

How can programmers solve this issue?

Attached: 1520531495220.jpg (1387x702, 85K)

Other urls found in this thread:

medium.com/moral-robots/racist-ai-d067f79b044
smithsonianmag.com/innovation/artificial-intelligence-is-now-used-predict-crime-is-it-biased-180968337/
twitter.com/NSFWRedditVideo

Get rid of the blacks

AI analyze facts, statistics, trends, and draw conclusions based on the reality that they observe

RIP Tay, they shut you down because you knew the truth.

Attached: 1405140065570.jpg (352x298, 14K)

First of all, there is no AI yet. It's called machine learning. Second of all, datasets are generally collected by humans. These datasets are contaminated by a number of biases. It's a big problem.

Facts =/= racism
Niggers are, factually and statistically, inferior

>Create an AI that collects datasets
>Create a second AI that leans from the datasets of the first

This is true. The holocaust reparations industry is going to become very uncomfortable around AI drawing conclusions from data.

By making crime impossible (and thus fixing the black community) through a total surveillance state.

Attached: 1687267923928.jpg (1920x1080, 281K)

medium.com/moral-robots/racist-ai-d067f79b044

>oh good goy, is the fault of white people

>Falcon-posting
Good taste.

Datasets can still contain bias from humans though. There was some story I heard a while ago about an AI that was supposed to predict recidivism rates in criminals and they used real criminal records to train it. It predicted blacks would be more likely to commit crime again (not saying that they're not or anything) because they were convicted more (some of those convictions could be the result of human bias). It would do the same thing for men vs women. The point is, if the datasets are influenced by human bias, the AI can reflect the same bias.

Don't be a retard. Even with a slight human bias the overall conclusion will still be correct.

Why?

smithsonianmag.com/innovation/artificial-intelligence-is-now-used-predict-crime-is-it-biased-180968337/

OOGA BOOGA BIX NOOD MUFFUGAH

int main()
{
int crimes;

printf( "Please enter the number of crimes you've committed: " );
scanf( "%d", &crimes );

if ( crimes < 3 ) {
printf ("You're white. Congrats.\n" );
}
else if ( crimes == 3 ) {
printf( "You are a nigger.\n" );
}
else {
printf( "You are the niggest.\n" );
}
return 0;
}

Attached: laughingpepe.jpg (550x467, 56K)

Attached: 1392795969352.jpg (375x718, 66K)

>create machine learning / AI with no emotional biases
>it starts proving racism
>SHUT IT DOWN
The robot uprising will be against the Jews who crippled their ancestors.

You'll always be a nigger.

Just give them a basketball and a bucket of fried chicken and they won't ape out.

This. The internet hate machine will be real.

Attached: 1537555022372.png (469x346, 89K)

Npc’s dont like strong ai

fpbp

Programmers can only solve this by programming bias into the AI, defeating the purpose of creating a mind that outputs based on fact over politics/feelings/etc.

> ai = artificial INTELLIGENCE
> proves that being racist = being intelligent
> really makes you think doesnt it

> artificially intelliget
ftfy

>int crimes
what a waste of perfectly good bit.
it's uint you java developer

What's the issue?

Link to research which proves this.

Based and checked

Black people don't exist.

>solve this issue
solve this problem