Editorial Contribution by Sally Writes
At a time when diversity is the watchword and everyone from politicians to sports stars is reminding us that Black Lives Matter, one industry remains stubbornly white and middle class. In the US, fewer than 10 percent of Computer Science graduates are from ethnic minorities. And of those who do graduate, attrition rates are significant. These are damning statistics. Technology is reshaping the world that we live in, yet the people responsible for writing the software are not representative of society as a whole.
Making diversity a priority
It’s a problem that has existed for years. Attending a tech conference a decade ago, you couldn’t help noticing that practically everyone was white and male. Today, the gender imbalance has reduced, although the men still outnumber the women. But analysis by the Kapor Center shows that between 2014 and 2020, black representation in America’s top tech companies increased by only one percent.
So why so slow? It’s got nothing to do with inherent racism in the industry, but is more down to inertia. The drivers are simply not there to make change happen. Putting that right is down to all of us. Suppose, when selecting an IT product developer, its diversity policy was a deciding factor in the approval process. That might sound fanciful, but if people really understood the potential consequences of parts of society being underrepresented in tech, they would take it a whole lot more seriously.
Racial disparities lead to unconscious bias
When one part of the community is underrepresented in the technology ecosystem that underpins our world, strange things can start to happen. Remember the furor a couple of years ago when Facebook was caught selling discriminatory ads that excluded minorities from seeing certain properties, jobs and financial products on the market? Then there are the AI algorithms that are biased such that your insurance premium quote will increase based on the color of your skin.
It seems incredible that fundamental blunders like these can still be happening in today’s world. Yet this is the same world that comes together for the BLM movement. It is a world in which the public is so quick to name and shame people for the smallest racial slur if it is committed by an empty headed celebrity on a reality TV show.
What is being done?
The first step towards addressing a problem is acknowledging it exists. The work of organizations like the Kapor Center has been invaluable in identifying just how serious bias can be in today’s AI systems. It’s acknowledged that the most effective way of preventing bias is by having as diverse a group of people as possible involved in the system’s creation.
There is talk of legislative measures coming in to address bias in AI, but that is attacking the symptom, not the disease. It is down to tech businesses to take diversity, and programs like apprenticeships, seriously. And they will only do that when customer demands mean ignoring it has an impact on the bottom line. We all use technology, so the future is in our hands.
Photo credit: Kevin Ku