Artificial Intelligence and Interpretations of Race
Reading time: 4 minutes
Editor's Note: British English. Some words in British English use "s" where "z" is used in American English.
Have you ever been stuck at an airport security machine because the camera wouldn’t recognise your face? Ever been unable to unlock your phone with your face? Ever had to wave your hand under a dryer several times before resorting to tissue? How would you feel about a computer deciding you are more likely to commit a crime because of your skin colour?
The terms ‘technology’ and ‘racism’ rarely intersect in our minds because we view technology as objective. It has no thoughts or preconceived biases . But as we have progressed in our development of Artificial Intelligence (AI), issues have arisen based on how it interprets different human beings.
What is AI?
Simply put, AI is a set of algorithms which makes decisions and automates tasks. It translates for Alexa, helps doctors analyse medical images, controls robotics in factories, and so much more. No one really knows how it works, not even those that create it. But the development of it is through training bots. Creating a set of them, giving them instructions/tasks, training them, and deleting those who are furthest away from the desired outcome to then train the ones who got it relatively right, and repeating the process again.
Whats the concern?
Faces (including your own) are collated translated into mathematical formulas which are added to a database. The AI can then identify faces from the data and search the database to find who you are, whether this be through security cameras, phone software etc.
The issue is that we are currently in a global AI rush to develop and implement the best AI in our societies in order to make important decisions such as whether someone can enter a country or what health insurance someone should receive. If the computers are making more mistakes with people of colour then future societies will be set up to automatically discriminate against such people, and potentially put them in harms way.
How has this happened?
The algorithms are only as good as the data inputted, so if the faces being used to practice AI are usually white, then the AI will only be able to accurately identify white faces. An article from Nature found that 45% of the most used image database in computer vision comes from the US. China and India, accounting for 36% of the world population, represents just 3% of data in the ImageNet dataset.
But will this actually affect my life?
Yes. This may sound very ‘Black Mirror-ish’ but in 2014 China has made their new social credit surveillance programme public knowledge with the intention of rolling it out by 2020. It uses a range of tech which I am not qualified to explain to rank people’s behaviour, identifying and logging actions such as bad driving, smoking in non-smoking zones, buying too many video games and posting fake news online. These scores are collated to determine where people can shop, travel, work etc.
But this isn’t all, Google’s image-recognition system failed to differentiate between black people from gorillas, there is a rising fear that self-driving cars will be more dangerous for black pedestrians, and Kodak’s colour film reduced dark skin to pitch-black smudges.
As AI becomes more engrained into the runnings of our everyday lives, it is important that computer scientists and social scientists work side by side to ensure that this wave of new tech is beneficial to all people are harmful to none.
Editor’s Words: This article gives us an underlying emphases on the need for black women in these tech spaces.
About the Author: Yes, she's another black girl with a blog. Cilla Hope writes about whatever angers or infatuates her. She’s a Junior Planner for a tech comms agency. Cofounder of TheMoveHub.com. A wine, steak & dark chocolate addict. Song rinser. Travels to wherever the suns out. Afro hair enthusiast.
Thank you for reading this post. We are an independent company and rely on readers like you to help support us! If you want to continue to support Your Corporate Black Girl, you can do so in a few ways:
Subscribe to us! Yes, it's that simple and easy!
Team Your Corporate Black Girl
MORE POSTS LIKE THIS