Sep 17, 2017 4 min read

My mother read the Bible daily, attended church twice a week and taught me that “the N word” was never to be used, no matter how many times I heard it thrown at the five or six African-American kids I attended school with in our small town.

But Mom had her ungodly foibles. For example, upon seeing this or that popular entertainer on TV, she would often say, with a straight face, “He’s a Jew. He doesn’t look like one, but he is.”

“So what?” I’d ask every time she made this observation. “What difference does it make?” As far as I can recall, she never gave me any kind of explanation, let alone a satisfactory one. But she had no hidden agenda, no ulterior motives. It was, to her, simply an observation.

I was reminded of Mom’s little quirk this week when I read the results of a Stanford University study undertaken to determine how well facial-recognition technology could identify people’s sexual orientation based on their faces alone. Researchers Michal Kosinski and Yilun Wang gathered 35,000 pictures of self-identified gay and heterosexual men and women from a public dating website, then ran them by an algorithm that noted differences, however subtle, in their features. Later, shown randomly selected faces and asked to guess whether they belonged to gays or heterosexuals, the algorithm responded correctly – 81 percent of the time for men, 71 percent of the time for women.

Kosinski and Wang wrote in their findings that because facial recognition technology is increasingly used by government and other agencies, “our findings expose a threat to the privacy and safety of gay men and women.”

Two LGBTQ advocacy groups, however, almost immediately took issue with this conclusion, calling the research “flawed” and “junk science.”

“Imagine for a moment the potential consequences if this flawed research were used to support a brutal regime’s efforts to identify and/or persecute people they believed to be gay,” said Ashland Johnson, director of public education and research for the Human Rights Commission advocacy group.

My initial reaction to the release of the Stanford findings was remarkably similar: “The next step will be to round up gays, make them sew pink stars on their clothing and force them into ghettos,” I thought. That’s an extreme reaction, to be sure – but, unfortunately, one not entirely beyond the pale. After some thought, however, I came down squarely on the side of the researchers.

“We did not build a privacy-invading tool,” Kosinski and Wang wrote in response to criticism. “We studied existing technologies, already widely used by companies and governments, to see whether they present a risk to the privacy of LGBTQ individuals. We were terrified to find that they do.”

And terrified we all should be.

We live in an age when mega-companies use algorithms to feed us information that reinforces what is, in many cases, a skewed worldview in which only those who think as we do are included in our increasingly shrinking circle of friends. If the men in your family grow beards, they might be branded Islamist terrorists. And now, at a time when you may be required to have your mugshot run through a national database before you can order a wedding cake.

The problem isn’t really with Kosinski and Wang’s findings. The problem is with an unthinking acceptance of technology, the timeworn “because we can do this, we should” approach to technical development. Our social and ethical structures aren’t keeping up with the bits and bytes churned out by the four or five tech giants that are already shaping how we shop, learn and connect. We should use Kosinski and Wang’s findings to make this type of profiling unacceptable and to start discussions about the right way to use technology.

Funny, we don’t look paranoid. But we are.