A Lot of Tech’s Big Problems Start With Who’s Building It
This article was originally published on 07/15/2019
There isn’t an area of your life that technology doesn’t impact. Although people tend to think of digital technology as something that stands alone, the reality is it didn’t just pop up out of nowhere. Generally, like the internet, tech can be used to give people access to things they wouldn’t otherwise have, like the ability to locate new information through search engines. While doing so, it can also reinforce social inequalities without even trying. For instance, how Google’s search engine was found to quietly reinforce racism.
While revelations around tech and its impacts on our society are a big part of our public conversation now, they aren’t too surprising since tech is an industry that doesn’t reflect the diversity of the people who use the products companies build.
Tech companies first started releasing diversity reports in 2014, and since then they’ve proven what most people already knew: That it’s a white, male dominated industry. According to the U.S. Equal Employment Opportunity Commission, about 83 percent of tech’s executives are white and at the non-executive level the picture gets even grimmer. A 2018 report by Reveal found that in 2016, ten large tech companies in Silicon Valley didn’t have a single Black woman in their ranks, while 3 of those companies didn’t have a Black employee at all. Recently, Google and Facebook released their diversity reports, showing Black representation in their workforces at 3.3 and 1.5 percent, with both companies still primarily dominated by white men in technical and leadership roles.
When tech is built primarily by white men, it begins to reflect whiteness — whether intentional or not, and its impacts are the same.
Currently, debates around facial recognition are consuming the country, as some cities seek to place temporary moratoriums and digital advocates calling to ban the technology in its entirety. One of the biggest issues surrounding facial recognition is both its inaccuracies and social impact. If you’re not a white man, chances are facial recognition isn’t going to do a very good job of recognizing you, and if you’re a Black woman the chances are even thinner.
This year, researchers from M.I.T.’s Media Lab found that Amazon’s Rekognition had greater errors in trying to recognize darker-skinned women. While testifying at a House hearing, researcher Joy Buolamwini — who is also the founder of the Algorithmic Justice League — told Congress about her experiences .
“In one test, Amazon Rekognition even failed on the face of Oprah Winfrey labeling her male. Personally, I’ve had to resort literally wearing a white mask to have my face detected by some of this technology,” Buolamwini said.
Black women are notoriously underrepresented in tech, so it should come as no surprise that the technology being built doesn’t serve them in a way that’s fair or even useful. Beyond being underrepresented in tech, Black women’s ideas don’t get funded at even half the rate of their white counterparts. According to ProjectDiane, Black women-led startups have received less than one percent of the total tech venture funding raised since 2009.
While tech’s diversity problems are vast, that’s only the first step in addressing the impact technology has on marginalized communities. The reality is that tech’s problems are reflective of structural issues extending beyond the individual. For example, Twitter can hire more Black people, but will that stop the company from allowing white supremacists to remain on the site because it fears conservative backlash?
Again, look at facial recognition as an example. Sure, the programs could potentially be taught to better “see” Black people, but that doesn’t erase the risks of widespread surveillance that makes so many people uneasy. It’s a warranted fear, given that the United States government has historically surveilled communities it found to be a threat, whether that was political positions such as anti-war efforts, but also based along with identity, like the continued surveillance of Black communities. If tech actually acknowledged history and listened to communities of color, facial recognition — and many other technologies — might not even exist. Just because we have the ability to create something doesn’t mean that we need to.
Tech by itself isn’t inherently oppressive, but one of the industry’s biggest strengths is that it has been able to mask itself as largely benevolent. Companies like Amazon claim they cannot predict the use of technology that they put into the world, despite warnings from communities who are directly impacted.
Part of the solution to tech’s problems is having more than just white men involved with it. People of color need to be careful not to replicate some of the inequalities that tech is presently reinforcing.