A few decades back Hitler and his cronies experienced with measuring human heads and determine who was worthy to live. A World War was fought to eradicate such researches, but now it seems that such scientific activities are back in full swing.
In a paper called “Deep neural networks are more accurate than humans at detecting sexual orientation from facial images” Stanford researchers are claiming that a trained AI was able to classify sexual orientation way better than human “judges” would. According to the paper this advance is actually done to avert classification (?) because
“given that companies and governments are increasingly using computer vision algorithms to detect people’s intimate traits, our findings expose a threat to the privacy and safety of gay men and women”.
So, in other words someone created a weapon against humans in order to point out to the possibility of it. Yea, got it.
Now, the research itself was centered around feeding the AI a dataset of more than 35.000 faces, designating gender-typical facial elements and then seek out the gender-atypical facial morphology in order to ID a gay. How they knew about the 35.000 people’s sexual orientation is not known, but if there was actually such a dataset, it raises serious questions in itself.
The research paper is seeking to find a modern scientific base for the so-called “prenatal hormone theory (PHT) of sexual orientation”, which holds that prenatal exposure to certain hormones influences the sexual orientation that emerges later in an adult.
Without diving too deep into this bizarre and gut wrenching story, I have to tell you that I am quite surprised that in the contemporary world scientists are actually getting paid for penning papers like this.
Recently it seems that there is a rush amongst scientists to develop even greater threats to the mankind in the name of pursuing “the artificial intelligence that will solve everything”.
Even though it seems that in this particular case they thrown the baby out with the bathwater…