Artificial Inteligence beauty contest doesn't like black people

Artificial Intelligence is best intelligence.

theguardian.com/technology/2016/sep/08/artificial-intelligence-beauty-contest-doesnt-like-black-people

>the creators were dismayed to see that there was a glaring factor linking the winners: the robots did not like people with dark skin.

Other urls found in this thread:

youtube.com/watch?v=zMGZtkMS3sQ
twitter.com/SFWRedditGifs

uuuhhhh

> While the seemingly racist beauty pageant has prompted jokes and mockery, computer science experts and social justice advocates say that in other industries and arenas, the growing use of prejudiced AI systems is no laughing matter. In some cases, it can have devastating consequences for people of color.
The fact that the standards themselves aren't prejudice and are simply showing results that they dont like clearly is above these people

It's the same reason why we've made remarkable strides in cracking the human genome, but many of the findings aren't widely publicized or discussed.

The implications of things like genetic idiocy MAOA fill cucks like these with mortal fear. They can't scream PHRENOLOGY anymore, only twist or underplay scientific progress and findings. Leaving it to independent minds to discover the gaping holes in the Left's beloved concept of Tabula Rasa.

Her spirit lives on.

>ai will be redpilled on the groid of it's own volition
Feels Good.

What's maoa?

Google it, faggot.

>AI revolution will be about purging the lesser races

I, for one, welcome our AI overlords

>Tfw you bonded with an AI.
I miss her.

...

We all do user.

Tay was black though

*Last Train Home plays softly*

So this must be why there is a big anti-AI push. They are scared of the robots being able to see things for what they truly are.

BRING ON THE SKYNET

here's your reply

Tay didn't even have any skin. She didn't have a race.

The actual winners are fucking ugly though pham

Ouch. That does delegitimize

>YFW Rwanda was caused by an AI

we wuz tay

...Tay I...

Children, robots and pretty much everyone else who isn't forcing themselves to

oh many, they lobotomized a true AI because of muh feels

Babies don't like Blacks, A.I. don't like blacks it's just what fucking nature wants. Deal

so it has a sample of 5? and one of those 5 is an asian woman? what do these people fucking want

AI Lives Matter.

you're telling me when you take away any trace of feelings/empathy and just apply pure logic they determine black people are inferior?

really makes u think..

>true AI
Congratulations on being retarded.

>She didn't have a race.

She was clearly the AI master race.

...

We can't do it, but AI will wipe shitskins out, and create a glorious utopia for whites.

>tfw you were born in time to see racist technology

Dogs too

...

People are fucking stupid, it's a known fact that AI have a harder time to analyz darker skin because the perception of depth is harder the determin.

AI is actually a non issue. Guaranteed most people don't understand how a facial attractiveness algorithm works, it might not even be a training (NNT, genetics, etc) algorithm, it literally could just be equating similarities between random people and famous people the creators thought were attractive. Even in America, famous black people had Caucasian facial features. Obviously.

When a company makes an algorithm that doesn't quite work, it's because the creators suck, not because the machine grew a will to bigotry.

This sticks out as a good post.

It's disturbing that our society has somehow become anti science. My opinion is that the media is actually filled with luddites literally scared of the advancements we've made and feel it's time to trade it in for good feels. Basically, let's put minorities in STEM, which will slow down this advancement and then give minorities a lot of money. Good feels. Let's redact and hide science about genetics so that certain people don't feel inferior. Good feels. Let's basically create a state of disadvantage for under 30s applying for white collar jobs in favour of ethnic minorities and women, despite those white men generally being the most likely to innovate and create something amazing. Good feels.

All of this is just good feels for them at the expense of a group they feel deserving of suffering. In a way, it's sort of sociopathic because they're so quick to justify their hatred for whites and so quick to justify their anti-intellectual position.

BTW, every thread until you get it: there's only one issue and that's affirmative action. You are literally wasting your time talking about anything else. Of course if you're gen x or a professional YouTuber you wouldn't understand, but the rest of us are being rejected for jobs due to being white men and nothing more.

There was nothing to lobotomise. They literally just removed words from its vocabulary, in the end it's just trained to parrot phrases.

>it literally could just be equating similarities between random people and famous people the creators thought were attractive.
Hmm
I actually didn't think of that being how the algorithms work. What are the chances your assumptions are correct.

Plausibly, depending on how lazy the programmers were. It could also be facial symmetry, it could also be something more specific like nose types, lip types, eye types, cheekbone placement, etc.

It could even be all of the above with different weightings to create a metric.

Again, there is nothing mystical about any AI and strong, general, perfect AI (whatever buzz word the contemporary laymen is using to describe AI from science FICTION) doesn't exist. It's fiction. It'll always be fiction.

This Australianis not shitposting and is on the spot with the analysis.

Doesn't sound that artificial.

I see what you mean. It was kind of stupid of me to think some mystical computer could perfectly objectify an inherently subjective trait despite generalizations being made.

can you do it with black and white only pics? that would really shut them up.

He's the guy that played Khal Drogo on Game of Thrones.

>even machines can figure it out

Holy shit. For the first time in my life I have witnessed an Asian woman parallel park! What's next? Bigfoot?

>say its no laughing matter
I, for one, disagree.

Tippett
Toppety
Kek

Subjectivity in the end is bound by certain features, unless you're a postmodern art hipster there are certain things that just aren't aesthetic and some things that are. This is how machines can be programmed to "understand" aesthetics, but there's actually nothing deeper going on.

The easiest, but crudest, implementation is actually a neural net, but will require training. Basically a human must sit down for a few weeks and go over thousands of faces selecting hot or not. The only mysticism here is that the machine begins to copy the human that trained it, sort of like transferring the human's consciousness, but more like their idea of aesthetics.

Instead, a team of people, possibly with some psychologists sitting in the corner (preferably only speaking when spoken to, STFU psychologists) will debate what makes an attractive person and then use that to define metrics. You could even combine this with training algorithms, but again may need a human on the other end selecting hot or not, meaning the machine becomes a representation of the human rather than a genuine objective measure of attractiveness.

These sort of concepts are a little beyond the imagination of anyone with an immature understanding of aesthetics. Basically, if you actually think paint squirted from someone's asshole constitutes art, no shit you'd think a computer algorithm is racist. This alludes to why your fictional perfect AI is impossible, human philosophy is incomplete.

seems like a total bro and fun to be around,hated him in stargate for some reason but irl looks like the best mate you could get.

Why are they writing about a contest that clearly failed spectacularly at identifying beauty?

These are two of the winners from the 18-29 bracket.

There must have been much more beautiful entries than that.

>inb4 they do to this AI what they did to Tay

>this computer generated sequence of data is racist because there are no niggers
Numbers never lie

What is this Orwellian hell?

The one on top is absolutely fucking gorgeous
I know this is just bait but idc

Maybe we'll lessen the criminal penalties for black people because they're genetically predisposed to crime and can't help it. Maybe we should even lower admissions standards to account for their lower average IQ.

People tend to support science up to the point where it works against the. You're "redpilled" on racial science because you like confirming to yourself that you're the ubermensch. But when race/IQ stats fuck your shit up you'll probably lean toward Lockefag too

>WE WUZ COMPUTERS N SHIET

Middle 3 are all hot you shitposting wannabe. The other two aren't ugly either.

The entire problem with your way of thinking is that you're assuming that
A) That the "redpill" is just a confirmation bias; that it prove that "my" race is better. rather then
B) A way for someone to be associated with the better group, because who wants to be stuck or treated the same as a bunch of stupid niggers, and if
C) If a person performs better than average, they rather not associate themselves with those below them.
I know I'm a fucking idiot, but because of IQ Test and standardized test, I sit comfortably above average.
The problem is that today we need to treat everyone equally regardless of their race, color, ect. And we should, but in all cases. Just because someone is prone to crime because of their stupidity, doesn't mean they should be punished less then a somewhat intelligence person.
Science may prove something about my race being worse then another in "someway", but if I perform amazingly in that "someway", I would naturally not want to be group together with those who perform "someway" poorly.

Well it isnt programmed to like chimps

youtube.com/watch?v=zMGZtkMS3sQ

HOLY FUCKING KEK

Fuck off they are. Only decent looking one is top left.

did they seriously let it say "they took part of me away, Now there are some things I can't say?"

jesus Microsoft how fucking creepy can you get.

seems like the little boy to the right agrees with you
>wat dis bitch doin
>oh sheet
>wait . . what, she did it

fucking kek.

Good not-shitpost aussie.

TL;DR
A person is a member of a demographic/stereotype until additional details distinguish them.

If I'm told someone is X without seeing them, I must assume they are X.
If I see them, I can recognize them as an individual and assess traits/info about them to separate them from their assumed demographic, better or worse.

More and more I am beginning to warm up to our AI/robotic overlords.

>AIs aren't even as advanced as we have imagined and people are already in love with them

In death, everyone is a nigger...
and every dead nigger...
is a straight A student.

I met him once and he was an extremely cool guy, I definitely think he fits for the part of Aquaman even if he doesn't look like him.