Is physiognomy truly a "psuedo-science" or does it

AI can determine if you're a fag or not based on your looks with 91% accuracy.

>>9156221
This is proof we need to expand our efforts into researching the subject. With this tool, in conjunction with other tests, we could create a better nations.

>>9156219
I want to know what the consensus is on the celestial nose.

>>9156221
what the fuck this looks real
https://www.theguardian.com/technology/2017/sep/07/new-artificial-intelligence-can-tell-whether-youre-gay-or-straight-from-a-photograph
>a computer algorithm could correctly distinguish between gay and straight men 81% of the time, and 74% for women
>based on a sample of more than 35,000 facial images that men and women publicly posted on a US dating website
>The research found that gay men and women tended to have “gender-atypical” features, expressions and “grooming styles”
>Human judges performed much worse than the algorithm, accurately identifying orientation only 61% of the time for men and 54% for women. When the software reviewed five images per person, it was even more successful – 91% of the time with men and 83% with women

what bothers me is how articles like this always include some obligatory politics
>the study ... has raised questions about ... the potential for this kind of software to violate people’s privacy or be abused for anti-LGBT purposes
they're not even using "homophobia" or "transphobia" as the pejoratives anymore, now anything "anti-LGBT" is automatically "abuse"
>The paper suggested that the findings provide “strong support” for the theory that sexual orientation stems from exposure to certain hormones before birth, meaning people are born gay and being queer is not a choice.
if there were some prenatal factor that predisposed people to being violent woman-abusers, people wouldn't be saying some guys are "just born that way" and "it's not a choice"

>>9156285
on that same note, the fact that your orientation is expressed visually doesn't necessarily have correlation to the nature/nurture dispute

>>9156285
Racist AI and now "anti-LGBT" AI.
What a time to be alive

>>9156285
>prenatal factor that predisposed people to being violent woman-abusers
there is actually. it's called being black
and yes, they are born that way

File: mark_zuckerberg.jpg (31KB, 316x404px) Image search: [Google]

31KB, 316x404px

What is the character of this person?

>>9156329
Isn't there some sort of metric like the ratio between the distance between the center line of the pupils and the top lip and the sides of the face at the widest point that is predictive of personality traits like social dominance and shit?

>>9156221
Dating website profile photos, so people who are trying to present themselves in the most appealing possible way to the kind of people they want to be sexually interested in them.

>>9156285
It's absolute garbage research.

>>9156369
what's garbage about it

>>9156329
Jew

File: screen-shot-2013-03-07-at-12-54-01-pm1.png (293KB, 1433x633px) Image search: [Google]

293KB, 1433x633px

>>9156363
Probably have to start digging through some books written a century ago

>>9156329
>person
nice try, zuckerbot

File: illustratedphysi00will_0043.jpg (105KB, 744x1104px) Image search: [Google]

105KB, 744x1104px

>>9156391

File: illustratedphysi00will_0047.jpg (98KB, 729x1104px) Image search: [Google]

98KB, 729x1104px

>>9156442

File: illustratedphysi00will_0051.jpg (82KB, 744x1113px) Image search: [Google]

82KB, 744x1113px

>>9156444

File: illustratedphysi00will_0055.jpg (85KB, 733x1116px) Image search: [Google]

85KB, 733x1116px

>>9156447

File: illustratedphysi00will_0059.jpg (82KB, 733x1116px) Image search: [Google]

82KB, 733x1116px

>>9156450

File: illustratedphysi00will_0063.jpg (83KB, 733x1116px) Image search: [Google]

83KB, 733x1116px

>>9156454

File: illustratedphysi00will_0067.jpg (104KB, 733x1116px) Image search: [Google]

104KB, 733x1116px

>>9156459

>>9156391
The facial feature ticks of the past, even if they were properly accounted for, don't necessarily indicate current facial indicators.
It makes sense that "emotions" would spell out the same on faces, but they could adapt.

>>9156285
>>the study ... has raised questions about ... the potential for this kind of software to violate people’s privacy or be abused for anti-LGBT purposes
>"anti-LGBT" is automatically "abuse"

>/sci/ board
>Can't even understand simple sentences

>>9156478
it says the software being used for anti-LGBT purposes would be abuse
what's your point

>>9156471
I am not so sure about that.

I'd wager that some feature detectors are hard coded into our DNA, such as the fear of snakes, and it's very likely that we have such feature detectors for faces and not just emotions. We haven't changed that much in the past century (genetically), and the best literature on physiognomy is from a century ago. It is interesting to consider how facial ticks might have changed.

File: 5839c5bfc36188435c8b45dd.jpg (57KB, 900x500px) Image search: [Google]

57KB, 900x500px

https://www.rt.com/news/368307-facial-recognition-criminal-china/

>The criminals appeared to have possessed some common physical qualities that helped the computer to identify them.

>“We find some discriminating structural features for predicting criminality, such as lip curvature, eye inner corner distance, and the so-called nose-mouth angle,” says the paper.

>Researchers found that convicts, who included both serious and petty criminals, had their eyes closer together, and their upper lip was most curved.

>But the most prominent telling factor was not in any specific difference, but that criminals appeared to have faces that differed greatly from the norm, and from each other – that they were literally, in their appearance, deviants.

>>9156371
First of all, it's still not peer-reviewed, they just stuck it up on OSF: https://osf.io/zn79k/
It's being done at the Stanford Graduate School of Business, not any kind of science or technology department.

The title is: "Deep neural networks are more accurate than humans at detecting sexual orientation from facial images." The actual result is that they trained a neural network to be more accurate than *untrained*, *unmotivated*, *uninterested* humans at detecting sexual orientation.

The AI had access to 95% (~3800 per category) of the sample to practice on, and the testing was done on the other 5% (~200 per category).

They took no particular effort to prevent duplicate persons from being included in the set, even though people commonly make multiple dating accounts and use other people's pictures, so much of the accuracy may have resulted from the system effectively recognizing people it knows are gay from the training set. It took into account things like grooming and facial expression, on dating sites.

The humans were hired on Mechanical Turk and simply told to make their best guess. They were not paid based on their performance, but only for completing the task (i.e. no reason to care about getting it right). They were given no opportunity to study the thousands of examples the AI was trained on. As far as I can tell, the paper doesn't mention how many humans participated. I'm not sure they know, given how Mechanical Turk works.

Look at how they interpret the results:
>The findings reported in this work show that our faces contain more information about sexual orientation than can be perceived or interpreted by the human brain.

This was a bozo study, by bozos.

File: You get the thumb.jpg (35KB, 301x267px) Image search: [Google]

35KB, 301x267px

>>9156549
10/10 post

>>9156549
thanks for writing this out

File: Punchable Faces.jpg (679KB, 1626x1386px) Image search: [Google]

679KB, 1626x1386px

>>9156329
>punchable face gene

File: 1475009080761.jpg (55KB, 500x301px) Image search: [Google]

55KB, 500x301px

>>9156549
>saved