“I think Artificial Intelligence will further broaden the gender gap,” I bluntly state.
My boyfriend reaches for his cup of coffee, a sign that he’s ready to listen intently. By this point, he’s used to me going off on feminist rants about living in a male-dominated society. I can’t recall how many times he has heard me say that “even the most “woke” people are still fighting to deconstruct the deeply ingrained patriarchal mechanisms within themselves.”
But, am I wrong?
Getting back on track, as an Engineer graduate (subtle flex), there is one thing I understand very well: data. So, allow me to back up my initial statement on why AI will widen the gender gap.
According to the World Economic Forum, women account for less than 25% of AI specialists, only 14% of the cloud computing workforce, and 20% of engineers. The AI research staff at Facebook is only 15% women, while at Google this figure is just 10%.
Those are really low percentages, considering that the AI market size is expected to grow 37% every year from 2023 to 2030.
For something growing exponentially fast, I think it is terrifying that this technology and these spaces are being designed without an inclusive perspective.
Reality’s Middle Name is “Harsh”
During South by Southwest several weeks ago, reality hit me square in the face. At three separate panels, accomplished women from various sectors echoed the same sentiment: the glaring absence of accessibility, databases, and representation for women. Though I was aware of these issues, their impact on the ongoing AI revolution hadn’t truly sunk in.
My first wake-up call was during a panel on how content creators use their platforms to advocate for social change, Michelle Hope, sexologist, activist, and Founder of Mixed Moxie LLC, said: “I think it’s (AI) being formed through a relatively patriarchal lens that is steeped in white dominant culture norms.”
“I was technically never supposed to belong here. These spaces weren’t actually built with us in mind,” she concluded.
While I knew Hope was primarily talking about her perspective as a sex activist, her sentiment transcends her field, echoing throughout various domains of society, including AI.
The 17+ Year Gap
For an AI model data is everything. If you don’t have enough data, what happens to your learning model? It will only go as far as the data you feed it.
Women are still years behind men, not only in the lack of representation in leadership roles, societal expectations, or economic opportunities but also in inadequate representation in databases.
Did you know it wasn’t until 1993 in the US that we were mandated to include men and women in clinical research? That means we have a 17-year gap of data.
During a panel on the top emerging AI trends, Sandy Carter, Senior Vice President, and Channel Chief at Unstoppable Domains, gave a clear example of the consequences women face due to the data gap.
“It (data) matters a lot, because now we have all of these new AI medical tools coming out, state of the art, amazing, that are working really great for the guys out there,” Carter said.
She continued by sharing the example of a liver disease prediction tool that can detect liver cancer with just your blood work. In 77% of the cases, it works really well for men, but for women, it fails in 44% of the cases, why? Not enough information.
How can there not be enough data??? While I’ve always been aware of gender disparities, I never imagined they could be reduced to 1’s and 0’s. The idea that gender could be encapsulated in code had never crossed my mind.
The truth is, AI isn’t going anywhere soon, and I foresee a persistent trend of these tools being primarily made by men, for men. Especially since data collection incurs costs, and those in charge (men) are reluctant to invest in closing the gap.
This brings me to the last panel I attended, on women in sports, where former WNBA champion Sue Bird and co-founder and chief content officer at TOGETHXR, Jessica Robertson, brought up the importance of data in the investment and growth of women’s sports.
Despite growing viewership and heightened public awareness of women’s sports on a global scale, they continue to receive significantly less media coverage. Research conducted in March 2021 by the University of Southern California and Purdue University revealed that women’s sports accounted for merely 5% of total TV sports coverage.
Even when men’s and women’s leagues are partners – as we see in the NBA and WNBA – they don’t collect and share the same statistics, because it would be too costly.
“That’s when you start to realize that the systems were never built for us to get the proper media deal,” Bird said. “They were never built for women’s sports to succeed in the way that men’s has been able to, which is really what equity is to me in these spaces.”
Mind you, even with data, the next problem arises: How do we share this information? Currently, only 10% of sports editors and 11.5% of sports reporters are women. If AI is leveraged correctly, media platforms could potentially analyze and recognize women’s sporting events, ensuring equal coverage and promoting their achievements.
The “What We Need” Manifesto
While women have opportunities, we can’t overlook the existing gap. We must work twice as hard and twice as fast to bridge it; otherwise with the rise of AI, equitable opportunities will remain distant for years to come.
We NEED to start paving our own way in AI.
We NEED to spend more money on research for women in every sector.
WE NEED MORE WOMEN DATABASES!
We NEED to work collectively to erase society’s stereotypes of women in STEM.
Women’s stories NEED to be taken into account when building new spaces. I am so tired of searching for a story that doesn’t exist – we NEED women storytellers.
We NEED men in high ranks advocating for women, as much as we NEED women to advocate for themselves.
We NEED to embrace AI as an opportunity to push for more visibility and research, and not as a tool to keep widening the gap.
And we really can’t be nonchalant about this. I am constantly reminded of a quote by Margaret Atwood from the Handmaid’s Tale that says: “Nothing changes instantaneously: in a gradually heating bathtub you’d be boiled to death before you knew it.”
We NEED to be more mindful of AI or we WILL be boiled to death.
