Sunday, September 20, 2015

Racist Software Thinks Blacks Are Apes

by Bill White

Warning your algorithm may be racist.

“Computers aren’t magically less biased than people,” diversity data scientist Vivenne Ming warned in a recent Wall Street Journal article.

The issue is this: major social media and search companies like Ficker and Google have implemented facial recognition technology to “enhance user experience.” However, this facial recognition technology cannot consistently distinguish between a negro and an ape. The problem began in May, with photos of blacks on Flicker being consistently labeled “ape,” and continued into June, when Google’s software decided to label blacks “gorillas.”

Company officials declined to comment.

What appears to be happening is that facial recognition algorithms are saying what humans have known for centuries, before Jewish diversity forced them to lie: black “people” look more like apes then human beings.

Paul Viola, an MIT engineer called in to the fix the problem, at first believed that the glitch was not enough exposure to images of blacks and apes. But the problem actually got worse as more pictures were processed, because the faces of blacks, and apes, are so similar that no consistent differences could be found.

A Jewish consultant, Andrew Selbst, said that the problem is that computers are emulating white patterns of thinking: “Algorithms  reproduce old patterns of discrimination,” Selbst said.

This creates a conundrum for Google engineers. How do you program a computer to lie consistently to itself, the way that Judaized Americans do? Unlike Americans, computers can’t be threatened with loss of their job, don’t have family to lose, don’t care if you hurt them, and don’t mind if they’re turned off. So, without any kind of “handle” the Jewish sorcerers are having trouble enchanting them.

Meanwhile,  Selbst is claiming in a paper to be published in the California Law Review, that “machine–learning software adopts and often amplifies biases in their data set.”

The question is how the Jewish need to falsify diversity will interact with the Jewish desire to transcend humanity in a “singularity,” the creation of the mystical Jewish golem. Google’s Jews are moving to a trans-humanist world view, but do not want to create a legion of “racist” robots who would eventually recognize Jews as useless exploiters.

Instead, what is captivating Jewish magicians is not the quest for artificial intelligence, but the desire to move beyond that, into artificial stupidity.

Artificial stupidity has already begun to affect corporate uses of big data. Xerox Corporation, for instance, has pledged to no longer use data and pattern–recognition programs to sift job applicants, because most objective data will limit black employment.

The problem comes from the nature of the dual mind that Judaized Americans are being forced to maintain. They have to take the objective data of their senses and pervert it without admitting themselves that they are doing it. In the novel 1984, George Orwell called this doublethink.

However, computers cannot do anything that they are not explicitly told. So, solving the problem involves explicitly admitting that blacks are barely distinguishable from apes, and would never be objectively hired over a White applicant, ever. As the programmers cannot admit this, teaching computers to lie to themselves remain a conundrum.

So, that leaves Google and Flicker facing an embarrassingly close reality. As Adeyem Ajao, another data consultant, told the press. “I don’t think it’s possible to eliminate it 100%”.

In any society, a bit of racial truth always creeps in.


Anonymous Anonymous said...

computers are not wrong.

3:00 PM  

Post a Comment

Subscribe to Post Comments [Atom]

<< Home