Artificial intelligence has gone from science-fiction fantasy to imminent component of the future in the space of just a few decades, with many of the fictitious genre’s predictions inching closer and closer to reality.
One such foreshadowing short story was Philip K. Dick’s The Minority Report, which was later made into a feature film directed by Steven Spielberg. In that tale, three mutant beings called “precogs” are able to foretell criminal behavior before it actually happens. The only problem is that they don’t always agree, creating an idea about what will happen based on what the “majority report” dictates. Placing faith in their educated guesses, the mutants tell law enforcement what a person will do — before that person knows themselves.
While mutant psychics are a far cry from artificial intelligence, there is an interesting parallel between this work of fiction and a new study from Chinese university Shanghai Jiao Tong, which suggests that a computer can prophesy whether or not someone will participate in criminal behavior in the future based on what they look like.
— Evan Kirstel (@evankirstel) November 16, 2016
The criteria for future criminals is based off of 1,856 faces that were fed into a computer, eventually leading to four artificial intelligence-based categorizations that the authors of the study claim can predict such aberrant behavior with “consistently well-performing” accuracy. According to a summary of their findings, “lip curvature, eye inner corner distance, and the so-called nose-mouth angle” are reliable indicators about what lies in the hearts of men.
“The variation among criminal faces is significantly greater than that of the non-criminal faces. The two manifolds consisting of criminal and non-criminal faces appear to be concentric, with the non-criminal manifold lying in the kernel with a smaller span, exhibiting a law of normality for faces of non-criminals. In other words, the faces of general law-biding public have a greater degree of resemblance compared with the faces of criminals, or criminals have a higher degree of dissimilarity in facial appearance than normal people.”
Still sound like word salad to you? To put it directly: People in the minority — outside of the homogenized group — are more likely to be a criminal according to artificial intelligence. Those in the majority, those non-criminal folk, are more likely to look the same. The computer, ostensibly free of racial bias, can simply sort people into criminal probability categories based on their physical appearance.
Such a statement may sound eerily familiar, particularly to the archaic study of phrenology. Even the study’s authors comment on it briefly in the abstract, praising the artificial intelligence’s capabilities “despite the historical controversy surrounding the topic.”
Long before artificial intelligence in the mid-19th century, both abolitionists and slave owners alike used the science as proof of the intellectual inferiority of those of African descent, reported The Guardian. Charles Caldwell, a doctor from Kentucky who imported the idea from Paris, argued that slaves were tamable creatures, whose brains were shaped in such a way that they required dominance to survive.
“Depend upon it my good friend, the Africans must have a master.”
While this particular development may disturb some, artificial intelligence itself isn’t all the doom and gloom of dystopian science fiction. In 2015, the United Nations pledged to end world poverty by 2030, a herculean task that would truly require an unprecedented level of cooperation and effort. Marshall Burke of Stanford University helped create an automated system which scans satellite images of the earth to find the world’s poorest communities and intuitively figure out what they need most. Such a technological advance could be essential for concentrating efforts to end poverty.
“We hope our data will be directly useful to governments around the world… to more effectively target their programs.”
Hopefully the companies at the forefront of artificial intelligence will be able to harness the power for good. In September, Facebook, Microsoft, Amazon, IBM, and Google created the Partnership on Artificial Intelligence to Benefit People and Society — a non-profit with the intention of “formulating best practices on the challenges and opportunities within the field.”
[Image via Colorado Springs Police Department/Getty Images]