How to make AI more diverse

LLast week, the New York Times published an article entitled “The People Behind the Dawn of the Modern Artificial Intelligence Movement,” highlighting 12 different people who helped make the modern AI movement a reality. Women are noticeably absent from the list.

The list has been rightly criticized for not including prominent female AI figures, including Dr. Fei-Fei Li. This also served as a springboard to talk about broader diversity issues in the AI ​​field. In 2021, only 21.3% of AI PhD graduates in North America were women. Less than one-third of AI jobs at U.S. tech companies are held by women, according to data shared with Charter by Rebellio Labs.

We recently spoke with Olga, an associate professor of computer vision in Princeton University's computer science department and co-founder of AI4ALL, a nonprofit organization dedicated to making AI more diverse and inclusive. – We spoke to Dr. Rusakowski about the lack of gender and racial diversity. love. Below are excerpts from our conversation. Edited for length and clarity.

What do you think is the root cause of the lack of diversity in computer science, especially AI?

Some of these include the correlation between socio-economic groups and access, lack of role models, and lack of visibility. Part of it has to do with implicit bias in hiring and hiring. And then there's the issue of fear of AI, which is different from other forms of computer science. There are issues like job loss and Hollywood's killer robots. If AI is being portrayed that way, why do students who don't understand that they are represented in the field and have to fight to get into the field fight? ?If they see that this is a field that is going to take away jobs and kill us all, why would they go to such trouble??

One of the things we're trying to do at AI4ALL is talk about all the ways that AI can actually change the world for the better. Many of these students are passionate about issues of climate, mental health, solving poverty, solving resource allocation and distribution, and accessibility in disaster relief and response. We discuss how AI is transforming all of these areas and helping solve the problems we see in our communities. That's the connection between what's going to happen in this field and how we can encourage more diverse groups of students to participate.

Given what's going on in the field right now, I definitely think we're under-utilizing the power of this technology to solve some of these problems. I've had a lot of different training, I'm working from different perspectives and backgrounds, I'm excited about solving this broad problem, and I'm really passionate about bringing my particular perspective, background, and training to the field. Not enough people are. field.

On your faculty page, 2015 survey A paper published in the journal Science found that women and black people are underrepresented in academic fields, where experts believe innate talent is the main requirement for success. How do you see that playing out in the field of AI?

There are many people who speak loudly of “brilliant” and “genius” progress. I don't know if anyone has studied where AI falls on the spectrum of perceived excellence, but given some news reports I suspect it's pretty high up there. I guess. [There’s this idea] It takes great brains to make progress, and these days for some students that means they have to go through training that starts in middle school. I feel like I'm already behind the curve in terms of getting into this field or making any kind of contribution.And it definitely contributes to people's driving [out]. We are implicitly discouraging them from even trying.

—

Charter recently published a research handbook, “Using AI in ways that advance worker dignity and inclusion.” This handbook provides a framework for bringing AI to the workplace so that more workers can benefit from the technology. Here, we introduce two of his frameworks that are particularly relevant to diversifying AI efforts.