Lack of Diversity in Tech Equals Bias in AI

With women accounting for just 20% of tech employees -- persons of color even less -- there is a clear lack of diversity in the tech industry. And while it has also been proven that diverse work teams promote better productivity and new ways of thinking, we still have a long way to go before the gap is bridged. This lack of diversity includes lower profitability and less innovation, but one of the most concerning problems is bias in Artificial Intelligence (AI).

According to researchers at AI Now Institute at New York University, “bias in AI systems reflects historical patterns of discrimination.” The report goes on to find that AI technology is being created mostly by wealthy white men and can, therefore, be harmful to people of color, gender minorities and other under-represented groups.

And we have seen this over the years! How many of these AI issues do you remember?

• Google Translate would automatically suggest words like “he” for male-dominated jobs and vice versa, when translating from a gender-neutral language like Spanish and Turkish.
• Software designed to warn people using Nikon cameras when the person they are photographing interpreted Asians as always blinking. 
• Word embedding characterized European American names as pleasant and African American ones as unpleasant.
• Google Photos’s automatic image labeling classified two teenagers as Gorillas back in 2015.
• AI-powered facial recognition systems had 34% more errors with dark-skinned females than light-skinned males.
• In 2016, Microsoft’s AI chatbot Tay was programmed to learn human behavior by interacting with other Twitter users. It was shut down after 16 hours when its tweets became a stream of sexist, pro-Hitler tweets.
• In May last year, it was claimed that Correctional Offender Management Profiling for Alternative Sanctions (Compas), was much more prone to mistakenly label black defendants as likely to reoffend and wrongly flagged them almost twice as much as white people (45% to 24%).

The biases in AI are introduced into software when the bias (even hidden) is held by the programmer or the environment around it that feeds the software information. What does this say about the tech industry? Obviously there is not enough diversity to make sure offensive biases do not enter into the software behind AI. 

It’s time to make a change

Changing the landscape of the tech industry begins with making educational opportunities widely available to the masses.

Studies show that students from wealthy backgrounds tend to outperform students from poorer homes on SAT exams that hold so much weight with many of the colleges and universities in the US. Why? Because wealthy parents can afford private tutoring to help their students earn top grades. They can also afford to have their children take the exams multiple times so that they have multiple chances to score well on these tests. And they can enroll their children in activities like music, sports, robotics and debate to help their children stand out from the pack. Some parents can afford to spend tens of thousands of dollars on private college counselors to coach the students through the entire application process and put in a good word to their contacts at elite schools. 

But how can students from lower socioeconomic backgrounds compete in a system so heavily stacked against them?

Unbiased admission processes + Income Share Agreements = Accessible First-Rate Education for All

Luckily, schools like Juilliard, School 42, and Holberton School (which I co-founded) have introduced blind or automated admissions processes. This type of application process allows for a more inclusive and diverse environment that offers everyone the ability to receive a first-rate education and to gain access to the future they have always wanted, but didn’t think they could achieve. And when it comes to those expensive tuition costs, many schools such as UC San Diego, Holberton, and others are beginning to go the way of the Income Share Agreement (ISA). Rather than paying a fortune and going into debt, graduates pay the school back after they find a job. If they don’t find a job, they pay nothing back to the school.

Everyone deserves a first-rate Education and hopefully a few schools, universities, and politicians (like Bernie Sanders and Alexandria Ocasio-Cortez) are working actively on that. But A.I. is already reshaping our world and our lives at a massive scale. We need more schools, institutions, universities, and companies to join us NOW, because the lack of diversity in tech is going to create a world of injustice if we don’t fix this asap.