(WXYZ) — It’s National Women’s month and this year's national theme is women who advocate for equity diversity and inclusion, recognizing those who are leading change in their fields.
At the University of Michigan, a group of women are leading change in the field of artificial intelligence, working to shape the next generation of computer sciences
I met with the impressive group of researchers to learn how their work is moving the needle on issues of diversity and equity AI.
Historically speaking, there have never been as many women working in the computer sciences as men. However, these women are working to change that.
"Do you see as many women in your classes as men?" I asked Rada Mihalcea, a professor and director of the AI Lab at U-M.
"No, I wish. There is a lack of diversity broadly in computer science, even more so in AI," Mihalcea said.
"It’s typically a very male dominated field," Sindhu Kutty, a lecturer in computer sciences, added. Her research focuses on AI.
"I think that’s a problem because we really need everyone at the table since we build technology that impacts everyone," Mihalcea said.
The AI Lab at U-M is made up of both brilliant men and women who all supported publishing this on their website: Women of Michigan AI.
"What kind of research do you do with artificial intelligence?" I asked.
"The big agenda of my lab is to work on technology that would have positive impact," Mihalcea said.
Other researchers working on AI include Elizabeth Bondi-Kelly, an assistant professor in computer science, and Lu Wang, an associate professor.
"I call my lab the ‘realize lab’ in hopes of realizing AI for social impact," Bondi-Kelly said.
"My goal is trying to improve their capability and also understanding what their limitations is," Wang said.
"One of the things I’ve been very interested in recently, has been how does bias affect the way that we make predictions," Kutty added.
This work will shape the next generation of computer sciences and it’s very important that it’s coming from women. A 2020 World Economic Forum report found that women make up only 26% of data and AI positions in the work force, even though women make up roughly 47% of the U.S. labor force.
Mihalcea tells me this isn’t because of a lack of interest from women, but because historically women get pushed out of computer science industries.
She says women were the first to work in computer sciences decades ago before it became a primarily male dominated field.
"History repeats itself. AI is really very lucrative, so again a lot of people want in, and some people are pushed out," Mihalcea said.
To the team of male and female AI researchers at Michigan, it’s imperative that women, and people of all backgrounds, work in AI as many biases can be traced back to a lack of diversity.
"There’s actually some studies that look at, if you look at any one specific model, you can actually look at what kind of biases that model has learned and that can be traced to what kind of data has been fed to it," Kutty said.
In the near future, it’s likely that most all adults and children alike will be using AI to search for answers and create their own work. The potential for both benefit and harm is great.
"I think this is the best time and maybe the worst time to work with AI," Wang said.
These researchers tell me in order for the most amount of people to benefit from artificial intelligence, this work, the work of women, is not just needed, but required.
"It’s mandatory that we have representation in order to design tools that will positively impact everyone and not just a segment of the population," Mihalcea said.
