Diversity In Artificial Intelligence Could Help Make It More Equitable

Of all computer science doctorates, only 1.6% were awarded to Black doctoral candidates.

In 2019, The Guardian cited a study by NYU that emphasized the critical need for diversity in artificial intelligence. “The urgency behind this issue is increasing as AI becomes increasingly integrated into society,” Danaë Metaxa, a Ph.D. candidate and a researcher at Stanford University focused on issues of internet and democracy, told the outlet. “Essentially, the lack of diversity in AI is concentrating an increasingly large amount of power and capital in the hands of a select subset of people.”

As we head into 2024, not much on that front has changed. In November, Wired talked to several prominent women in the artificial intelligence community about why they would not want a seat on the board of OpenAI following Sam Altman’s coup. Timnit Gebru, who made waves when Google dismissed her following a warning she issued regarding the company’s plans for AI, said that there was a better chance of her returning to Google than joining Altman’s board.

“It’s repulsive to me,” Gebru said. “I honestly think there’s more of a chance that I would go back to Google—I mean, they won’t have me and I won’t have them—than me going to OpenAI.”

In this subsection of artificial intelligence, the field of AI ethics, women in tech have found a measure of success, but their work in the field often puts them at odds with the white men who control the boards and companies in Silicon Valley. Meredith Whittaker, the president of Signal, an encrypted messaging app, says the problem is about giving people from diverse backgrounds the power to effect change instead of tokenizing their seats at the table.

“We’re not going to solve the issue—that AI is in the hands of concentrated capital at present—by simply hiring more diverse people to fulfill the incentives of concentrated capital,” Whittaker told Wired. “I worry about a discourse that focuses on diversity and then sets folks up in rooms with [expletive] Larry Summers without much power.”

Black people, in particular, have felt the brunt of the way artificial intelligence is used by the police, for example.

As BLACK ENTERPRISE previously covered, the city of Detroit was sued by a Black woman who was arrested while eight months pregnant because officers used a facial recognition program to tie her to a crime. And this is just one of many similar incidents.

In a November article for Esquire, Mitchell S. Jackson surmises that this is inescapable as the field of criminal justice insists on pushing to use artificial intelligence, even though the datasets those programs will use are filled with negative biases that will inevitably work against Black people.

Jackson writes, “AI in policing is being implemented into that already flawed system. It’s more dangerous to Black and brown people because the persistent lack of diversity in the STEM fields—from which AI comes—is apt to generate more built-in biases against people of color, the same people who are overpoliced and underprotected.”

He continued, “AI in policing is hella dangerous to my people because it operates on data—crime reports, arrest records, license plates, images—that is itself steeped in biases.”

According to a 2023 report by the Code.org Advocacy Coalition, only 78% of Black high school students had access to foundational computer science courses, compared to 89% of Asian high school students and 82% of white high school students. A 2022 survey from the Computing Research Association says that two-thirds of all computer science doctorates went to non-permanent U.S. residents for whom no ethnic background is available. Still, almost 19% of those degrees went to white doctoral candidates, and 10.1 % were awarded to Asian doctoral candidates. Only 1.6% were awarded to Black doctoral candidates, which illustrates why the diversity numbers in technology companies are abysmal.

Calvin Lawrence, the author of Hidden In White Sight, a book examining how artificial intelligence contributes to systemic racism, spoke to CNN about how the biases in AI are also a product of a lack of access. Lawrence explained that to get more Black people into the field, you have to present it as a path they can take.

“You certainly don’t have a lot of Black folks or data scientists participating in the process of deploying and designing AI solutions,” Lawrence said. “The only way you can get them to have seats at the table, you have to educate them.”