Few people understand the gender imbalance women in tech face better than Gabriela de Queiroz.
While she was promoted to chief data scientist in artificial intelligence at IBM last May, she’s a rarity in the world of AI. At Facebook and Google, respectively, just 15% and 10% of AI researchers are women, according to a study in 2020 by the AI Now Institute of New York University. Another study by the Navigating Broader Impacts of AI Research pointed out that biased predictions are mostly caused by imbalanced data but that the demographics of engineers also play a role, as reported by VentureBeat.
de Queiroz isn’t about to let those stats go unchecked. Since her promotion at IBM she’s been focused on building a new team of 10 people from scratch. She had a goal to hire a diverse team from the start. Her team is comprised of 90% women and together her team members speak a total of 10 different languages. Meanwhile, she is running her second non-profit that already reaches over 20,000 people. Called AI Inclusive, the NGO hosts events and workshops across the world to encourage and promote women and minorities in the artificial intelligence community.
“It’s very hard to reconcile everything,” said de Queiroz. But she’s motivated. She understands that if she doesn’t work on improving the gender imbalance in AI, it could hurt the industry–and women’s role in it– for years to come.
Worldwide revenues for the artificial intelligence market, which includes software, hardware, and services, are expected to grow 16.4% year over year in 2021 to $327.5 billion, according to International Data Corporation. By 2024, the market is expected to reach $554.3 billion, with a five-year annual growth rate of 17.5%, shows the report.
In the beginning
de Queiroz ‘passion for statistics came from the empowerment and decision-making she felt while handling data. She could implement and solve real problems. But she quickly realized she had no space to put it to use.
After graduating in Epidemiology at Fundação Oswaldo Cruz, in Rio de Janeiro, where she was born and raised, she moved from Brazil to the US in 2012 for a second master in statistics at California State University. Soon after she was pulled into Silicon Valley’s web working as a data scientist in fintech and adtech startups. But reality sank in.
“My team was all men. I wasn’t a manager yet, so I didn’t have the power to change the hiring part; I didn’t have anything, in fact, I didn’t have a voice there,” said de Queiroz. “It was very difficult, it was a moment when I had to stop being who I was and follow what I was told.”
But de Queiroz knew that even though her experience showed the opposite, there was a better alternative for a male-majority technology industry.
Alternative view
She came up with the idea for her first non-profit R-Ladies as soon as she moved to San Francisco in 2012. She was one of the few women among many men to participate in data science and machine learning meetups. She was learning a lot. But also struggling to feel part of the group.
So she organized her own meetup – an all-free-coding workshop dedicated to helping women and minorities learn the R statistical programming language – R-Ladies. The events started locally and then began to cross borders. Today there are 138 chapters in 44 countries and 39,000 members. R-Ladies’ website is full of stories like Zulemma Bazurto Blacio, who founded the R-Ladies’ chapter in Guayaquil in Ecuador after realizing how a community could help in her goal to democratize statistics learning.
Being part of a network also helps women to get their foot in the industry. Alison Hill was in academic research when she started to attend R-Ladies workshops in Portland in 2017. Today Hill is a senior data scientist part of de Queiroz’s team at IBM.
“I was feeling totally alone,” said Hill of her experience prior to joining R-Ladies. “I was trying to find a community of people that I could talk to, to discover new tools to just get support from each other. And then I was like, ‘oh, wow, somebody already came up with this wonderful idea.’”
Challenges
AI doesn’t just have a gender problem, however.
“Many of the discussions over bias in algorithms are still happening in tech hubs, such as in the US and China. But the whole world will be impacted, especially minorities,” said de Queiroz.
Several studies already show the impacts of the lack of diversity in the data science community – facial recognition models, for example, fail to recognize Black, Middle Eastern, and Latinx people more often than those with lighter skin, as reported by VentureBeat.
Seeing these reports and witnessing the Black Lives Matter protests, de Queiroz began to feel the itch of something new. In 2019, she stepped down from R-Ladies global and founded AI Inclusive, which works to organize events in the US and Brazil for minorities to get in the AI community. But the COVID-19 pandemic quickly changed the plan and de Queiroz and her team had to shift online every class and workshop. She focused on seeking partnerships with companies like DataQuest, FourthBrain and R For the Rest of Us, which led to helping create 2,000 scholarships across the world. And she had to squeeze in office hours in her schedule to help out those who couldn’t follow the online content to avoid anyone giving up.
“She’s one of the most inspirational people that I can think of in my world (of data science),” said Stephanie Hicks, a biostatistics assistant professor at Johns Hopkins University and a R-Ladies member.
Hicks is currently writing a children’s book on women in data science in which de Queiroz is one of the characters. “I want them (the children) to know that people like Gabi exist and then hopefully they feel inspired to do similar things that can really impact the fields of statistics and data science,” said Hicks.
Aline Arakaki, de Queiroz’s childhood friend, said the data scientist has always shared her knowledge with others. When de Queiroz was a teenager she started teaching math, physics and chemistry to help out students that struggled with their studies.
“Everything she does is because she believes in it,” said Arakaki. “Doesn’t matter if she’s in a leadership position or not, she has always been like this and today the more she conquers, the more proactive she is in ‘giving back’ to society.”
The way forward
Now leading a mainly female team of 30 people across IBM and AI Inclusive, she carefully balances her leadership skills with her sense of community. She’s there in every Zoom meeting, but only sometimes steps in to encourage everyone to voice their opinions.
“Before I started to work with her, I didn’t know the importance of creating an environment where women feel safe to talk and evolve,” said Marina Borges, part of the AI Inclusive team. Borges was inspired by de Queiroz when she founded a community for women at Salesforce where she works as a developer consultant.
de Queiroz said she doesn’t know if in 10 years she would have already moved on from AI Inclusive to a next adventure. But she can’t think about that just yet. Her goal is to seek more partnerships to offer more scholarships across the world.
“I don’t consider [diversity] an activism, but a statement,” said de Queiroz. “It’s part of my values to have a diverse team around me that represents the world and not just Silicon Valley.”
Leave a Reply