They say you can’t judge a book by its cover. At one point or another, we have all been guilty of this, stereotyping people based on their appearance both consciously and not. Well, what would you say if there was a machine that actually could judge a book by its cover, or the sexual orientation by its face, more rather.
Polish psychologist Michal Kosinski and co author Yilun Wang, released an academic paper surrounding a very controversial study that many have immediately labeled as taboo and has sparked heated debates. The two researchers built a program using artificial intelligence (AI) to scan over 30,000 images uploaded from a US dating website, designed to identify the sexual orientation by analyzing a portrait photo. According to Kosinski, the program would make their decisions based on differences in facial structure along with some other defining characteristics in the face. Well, the software didn’t lie. In fact, the results were surprisingly accurate and they were amazed. Of the 30,000 + photos scanned for testing, the artificial intelligence algorithms managed to correctly identify men who were gay 81% of the time and lesbian women 71%. The data also identified certain trends, including that gay men had larger foreheads than straight men, and that gay women had larger jaws and smaller foreheads compared to straight women.
The study proclaimed that biologically speaking, a gay man’s facial structure proved to be more effeminate and a lesbian’s more masculine. By what did they mean more feminine or masculine? According to this computer based science, the Stanford researcher said “Gay faces tended to be gender atypical. Gay men had narrower jaws and longer noses. Lesbians had a larger jaw.”
Why and how would this be the case? Researchers Kosinski and Wang concluded that the success of the AI findings followed a theory that existed long before the study called the Prenatal Hormone theory of sexual Orientation. According to theory, one’s sexuality is determined before they are born, based on hormonal exposure in the womb.
Basically, characterististics like the faintness of beard and mustache in gay males is linked to the under exposure of male hormones like testosterone which could cause a feminizing effect, meaning less hair.
More importantly, Kosinski claims that Artificial Intelligence algorithms will be able to measure intelligence, political orientation and a number of other characteristics of our personal lives that we keep to ourselves. This wasn’t the first time this controversial idea had been brought up. In the past, there were two computer scientists from China that had released a paper about AI programs and its ability to accurately categorize criminals when simply give a government issued photo ID. The study resulted in 90% accuracy.
As one can expect, this controversial ideology along with the study has incited media uproar. LGBQT groups deemed the study a threat to safety of those who rather keep their sexual orientation private for personal reasons. For many gay men and women, “coming out of the closet” could be a potential danger in communities that oppose it. Many researchers have already decried the project as physiognomy, a form of pseudoscience that emphasized the connection between personality traits and physical traits. Physiognomy and pseudoscience have an incredibly racist and messy history so many research groups don’t want to have anything to do with it.
Major advocating groups for the LGBQT community like the Human Rights Campaign and Glaad have labeled it “junk science” and “reckless findings”, and urging Stanford to distance themselves from the study if they knew what was good for them. Glaad’s chief digital officer announced his sentiments, stating that “at a time where minority groups are being targeted, these reckless findings could serve as weapon to harm both heterosexuals who are inaccurately outed, as well as gay and lesbian people who are in situations where coming out is dangerous.”
Ethics aside, the method applied in the study was met with much criticism as well because a number of factors including the fact that the majority of the photographs used in the study were disproportionately made up of caucasians. Communities pointed out that the study was too narrow since it lacked diversity. According to Kosinski, it was difficult to find minorities that were openly homosexual even though minorities actually are statistically the most openly gay persons.
Also many have noted that the research was not taken in a lab environment that was neutral and rather was uploaded by users, which symbolizes many limitations.
"This research isn't science or news, but it's a description of beauty standards on dating sites that ignores huge segments of the LGBTQ (lesbian, gay, bisexual, transgender and queer/questioning) community, including people of color, transgender people, older individuals, and other LGBTQ people who don't want to post photos on dating sites," said Jim Halloran, chief digital officer of Glaad, a media-monitoring body.
Critics claim that rather than hormonal exposure, fashion trends and cultural norms are at play when it comes to facial characteristics and appearance. They negated the early development hormonal exposure theory, classifying the scientific research as stereotyping and surface-level generalizations by equating one’s sexuality with appearance.
But Kosinski and fellow Stanford researchers roll their eyes at the campaigns and those countering their study. They feel that they don’t realize the underlying notion of this announced concept, that this is actually a warning sign in alignment with the ethics of LGBQT groups. “Going forward, there’s gonna be no privacy whatsoever,” Kosinski says. The researcher argues that the public is not aware of this yet, but many aspects of our inner lives are actually visible on our face and the age of privacy has come to an end. According to him, this form of technology is not as breakthrough as we may think. The polish researcher believes this computer intelligence has existed for years and is safely hidden by governments and top-tier scientists, unknown to the public. Kosinski claims that when first deciding to announce such a concept as controversial as this, he realized it was his calling to expose the potential dangers that could ensue. It was of moral obligation to release the research paper, since he felt certain that private companies and other industries out of our reach were already misusing facial detection technology in sneaky ways that are unethical by definition.
Moreover, the Stanford scientists are confused by the oppositional stance of LGBQT groups when the study itself fully supports the common notion that these same groups perpetuate regarding homosexuality being something you’re born with rather than a conditioned behavior. He says that technically it’s actually in allyship with the LGBQT since it’s scientific proof to anyone who believes that queers are “acting out” or “going through a phase”. Rather, it’s due to under or over exposure of male and female hormones and the proof is right there on their face.
The public eye, however, is eating this breaking news right up and the crazy concept that the Gaydar actually exists is spreading like wildfire. There’s been heated debates all over Reddit and other forums as to if it’s actually true and legitimate or if it’s all just a scam that lumps pre-conceived notions fueled by stereotypes of gay people from roles we see on TV. After all, Teen Wolf actor Charlie Carver is as manly as it gets, but he’s also gay as can be. Or what about that hottie from Prison Break, Wentworth Miller? Could you really determine that they’re totally gay just by looking at them?
News outlets picked up the story as soon as the paper released and Kosinski and Wang found themselves (well, their study) flaunting the cover of The Economist’s Sept 9 issue, with a detailed story on their findings. Other respected publications in the psychology department like the Journal of Personality and Social Psychology published their work and Vox published an in depth interview regarding the study.
So, the real question here is could the Gaydar really and truly exist? This coined term of the Gaydar, defined by Urban Dictionary as the gift/ability to detect homosexuality in other people- Now comes with a whole new meaning. Is it possible that we’ve come to an age where a computerized device actually has the ability to detect gayness? The reports from the Artificial Intelligence study were freakishly successful, after all.
Everyone has that gay friend… Had you not been told that they had a thing for folks of the same gender would you be able to tell? Could it be written on their face? We know that the archetypes of gay guys in movies and TV shows are men who strut around and exclaim “Fabulous!” every chance they get. For lesbians, they’re often portrayed as the “butch” types, sporting short hair and are of a larger build with squared shoulders. But is this even relevant to all queers or is this study sorely mistaken and just a product of societal norms that are merely just assumptions and “junk science” that does more harm than good.
You can show your pride as a queer person or a heterosexual displaying your allyship with the LGBT community by wearing one of these bold shirts that proclaim that love wins, no matter who you’re with- Because it shouldn’t matter. Some shirts of this LGBT collection flaunt the gay pride rainbow of colors, while others proclaim loud statements like “The world has bigger problems than boys who kiss boys and girls who kiss girls.” There’s a variety of shirts for queer men and women screenprinted with funny sayings like “Oh deer! I’m queer!" These fun shirts are sure to make the kind of statement that advocates for a world where your sexual orientation should not define who you are and that it’s totally okay to be gay.
Another thing is the crazy conspiracy behind the whole project. Kosinski’s message that private industries are able to detect our personal beliefs and sexual behaviors is a spooky idea that’s making a lot of people feel really uncomfortable. It’s got a “Big Brother” element to it that sounds as if it’s straight out of George Orwell’s book 1984. Given the fact that selfies are part of our everyday lives as we actively use Snapchat, Facebook and Twitter pretty much all the time, the idea that so much can be read simply from facial detection makes us second guess if we’re even safe posting photos on our favorite social media sites. Could a simple snapchat selfie you sent to a friend secretly being recorded and logged into a database that categorizes your religious beliefs or sexual orientation? If so, who is obtaining this information and what are they doing with it that we’re not aware about?
Konsinski has told the public that he
would love for these frightening discoveries to not be true. But he
believes that there’s a lot going on that we’re not aware of and he’s
determined to find out.
"If our findings are wrong, we
merely raised a false alarm," reads a statement made by Stanford
researcher. "However, if our results are correct, GLAAD and HRC
representatives’ knee-jerk dismissal of the scientific findings puts at
risk the very people for whom their organizations strive to advocate."