Where are the women in tech and AI?

When she was a computer science postdoc at MIT, el Kaliouby learned from a pioneer — scientist and inventor Rosalind Picard — and, together, they founded Affectiva, a company at the vanguard of using AI to understand emotion. She sees AI reshaping every company and, in the process, transforming the nature of work. Those who “can leverage AI” to do their jobs more effectively will ultimately win big.

In the 1970s, el Kaliouby notes, Steve Jobs and Bill Gates foresaw a world where everyone would have a personal computer. Now, out of a worldwide population of 8 billion, about 7 billion walk around with hand-held computers, and someday we may all have AI assistants. But, to make that work, el Kaliouby says, we’ll need to bring more diverse brains and perspectives into computing, because the algorithms that power AI reflect those who create them. “I’d love to see more women founders. I’d love to see more women investors,” she says. “But at the very least, I’d love to see more women around the table when these problems get designed and solved.”

For more than a century, women dominated computing. In the late 1800s, women “computers” staffed a Harvard astronomy lab, calculating data and analyzing images. In the middle of the 20th century, women worked at NASA on early IBM computers, laying the groundwork for space exploration. During World War II, Grace Hopper worked on the Mark I computer used in the Manhattan Project.

Even in the early 1980s, when Carla Brodley was falling in love with software engineering, there were still a lot of women in the field. In 1984, 37 percent of undergraduate computer science degrees were awarded to women.

But the degree wasn’t yet the golden ticket it is today. Now, computer science is the single most lucrative profession you can enter with a bachelor’s degree, according to the National Association of Colleges and Employers. Salaries start at nearly $87,000, compared with $60,000 for the average college graduate. Brodley — who would go on to become dean of Northeastern’s Khoury College of Computer Sciences and founding executive director of the Center for Inclusive Computing — made about $18,000 at her first job, a salary that equates to just under $49,000 in today’s dollars.

The 1980s would change everything. The PC market took off. Gates and Michael Dell got famous. The media began to depict computer scientists as geeky guys. Kids (often boys) started using family-room computers to play video games. Brodley was in school at McGill University in Montreal, planning to major in English, when she discovered coding. “I felt like: Oh my god, this is the best thing ever. I want to do it for fun.”

She called her mother and said she probably needed to change her major. After confirming her daughter would still graduate on time, her mother asked: “Is there anything that makes you forget to eat?” Brodley didn’t hesitate: “I was like: Yeah! When I’m coding, I forget to eat. And [my mom] said: It’s a no-brainer. You should do this.”

But just as Brodley was embarking on a career in computer science, during which she would become an editor of journals on machine learning and artificial intelligence, and serve on the Defense Advanced Research Projects Agency’s Information Science and Technology Board, lots of women started heading for the exits. Though they had constituted nearly 40 percent of computer science graduates in 1984, by the early 2000s, women made up fewer than 20 percent. It was hard to argue that women weren’t good at working with computers. Or that they couldn’t do math. This wasn’t biological; it was circumstantial. Something profound had changed.

“All of a sudden, the world of computer science burst open,” Brodley says. Some kids — mostly boys, mostly in wealthier ZIP codes — started to dabble in coding during high school, or take classes in the subject. Professors could get more lucrative jobs if they went to work for industry. And colleges, facing a surge of interest and not nearly enough professors to accommodate it, often whittled down the student population by using GPA cutoffs based on first-year grades, which privileged students who had done some coding in high school.

Growing up in the ‘90s and early 2000s, Brittany Greenfield loved tech. “We had a server in our house,” she remembers. “There is an early, early picture of me sitting on my dad’s lap with one of the first IBM laptops, which was the size of a desk.” She went to an all-girls high school and took the Advanced Placement exam in computer science, planning to major in the subject at Duke University. But when she got there, she says, only one or two other women were in the introductory class of about 25. “It was shocking to the system,” she remembers.

Greenfield calls herself a “Comp Sci dropout” — one who went on to start Boston-based cybersecurity company Wabbi in 2018, where she’s CEO. She also sits on the board of trustees of the Massachusetts Technology Leadership Council and uses that platform, among others, to promote opportunities for women in tech.

When parents tell her that their teenage daughter is great at art and sociology, she tells them, “You know what that makes for? A great UI/UX designer.” She’s concerned that nobody’s helping girls “draw together those threads.” And she worries that many people — perhaps without realizing it — still think of a techie as a guy who sits in his basement.

Men’s dominance in such a lucrative field has enormous ripple effects. For couples, it’s perfectly rational to prioritize a $200,000-a-year job over a $70,000-a-year job, relocating to accommodate the career trajectory of the spouse earning more. When a family member gets sick, or children are young, earnings can also dictate who takes time off. But those micro-level decisions end up reinforcing a society-wide inequality, one that many people believe we’ve moved beyond.

Brodley wants women to grab some of the money that’s out there in computer science, and she doesn’t want baby steps. “I don’t think that having a women-in-computing support group in your university is the solution, if you want to fix the structural changes,” she says. “I’m really tired of women finally getting to a field when the money is kind of gone.” Computer science offers the sort of salary, she adds, “that can pull a whole family out of poverty.”

To get women some of that money, Brodley has spent years rethinking the traditional path to a tech career. Starting in 2014, she helped build out a transitional master’s degree at Northeastern that lets students with any undergraduate major get into programming. The year-round Align master’s in computer science program takes between two and two-and-a-half years to complete, and has now produced more than 1,300 graduates, nearly half of whom are women.

In 2017, Pivotal Ventures, founded by Melinda French Gates, plowed money into the program to help it reach more people. French Gates earned a degree in computer science and economics in the mid-1980s and rose through the ranks at Microsoft, ultimately becoming the general manager of information products. “The time when she was doing her undergraduate degree was at that peak,” says Renee Wittemyer, senior director of program strategy at Pivotal. “She was surrounded by a much more diverse set of students.”

In recent years, French Gates has intensified her commitment to helping women get into tech. “We know,” she has said, “that women, even today, who make it through computer science undergraduate degrees and then go to work for one of the big companies — Google, Amazon, Facebook, Microsoft — they will tell you that if they join a team that’s all male, they’re just not as comfortable.”

After Pivotal helped fund Northeastern’s transitional master’s degree program, they kept talking to Carla Brodley about why computer science departments were failing to attract and retain undergraduate women. Plenty of research had been done on what worked, but no one was synthesizing it and implementing it in a scalable way.

Brodley knew that colleges needed to eschew GPA caps, which tend to privilege those who studied computer science in high school. She’d seen interdisciplinary majors — such as computer science and design — attract more women. Those with and without coding experience should be initially split up, and then reunited after a few semesters to avoid instructors drawing conclusions about who is smart (those with coding experience) and who isn’t (everyone else). And, ideally, coding should quickly be applied to real-life problems, such as global health, so students can easily see the applicability of their work.

Computer science departments also needed to become more accepting and reject a sink-or-swim ethos. At Northeastern, Brodley found, even if teaching assistants were brusque, men shrugged it off. “I did a focus group here and interviewed two men and two women — in each case, one white and one Black. And the two men both said: ‘The TAs are jerks, but I found the one that isn’t, or I go to my friends.’ And the two women said: ‘They make me feel like I don’t belong, and I’m thinking about dropping.’”

Once women started flowing into classes in greater numbers — 30 percent is an important threshold, Brodley believes — ”They felt confident and empowered.” The tone changed. And the drop/fail/withdraw rate fell from 25 percent overall (close to 50 percent for women) to 5 percent. After implementing a slew of changes, Brodley went on “an apology tour” to tell Northeastern students and their advisers that it was safe to take computer science, no matter who you were.

Pivotal wanted Brodley to scale up her program. In 2019, Pivotal became the first funder of The Center for Inclusive Computing, which subsequently attracted other funders and is now a $46 million endeavor. Its aim is to usher colleges around the country through many of the changes that Brodley made at Northeastern, increasing the number of underrepresented students in computer science by improving the undergraduate experience. The center now works closely with about 30 schools, including campuses in the University of Texas, University of California, and University of Illinois systems.

For Carolina Lopez-Treviño, choosing to major in computer science was intimidating. Students around her, it seemed, “had been coding since they were 12,” she says.

Lopez-Treviño is the New Ventures program technical lead at the Boston-based software company Toast, which, she says, is like being “a CTO of a tiny startup” inside a publicly-traded company. As someone who ascended to a position of leadership in her 20s, and now gets to build her own team, Lopez-Treviño works hard to make sure that team is diverse, in part because it makes the end product better. But, she’s learned, “You can’t just put up a job listing and say ‘I’m hiring!’ and hope to get a bunch of people of color and women — you have to go and find them.” She says she reaches out to potential candidates personally on LinkedIn.

Lopez-Treviño attended a science-and-engineering magnet school in San Antonio and ended up at a school she’d never heard about before high school: MIT. There, she visited a job fair her first year and realized no one wanted a biological engineer, which was what she had hoped to become. They wanted computer science majors. “And I was like: Whoa, wait a second. I’m not just at MIT for fun. I’m at MIT to have a successful career. I want to be able to support my family. I want to be able to do all this stuff,” she recalls. “So if I’m going to do that, I’m going to go all in. If they want computer scientists, I’ll be a computer scientist.”

Though women are a distinct minority in tech, Hispanic women are missing almost entirely. They constitute around 9 percent of the US population and 17 percent of all women in the workforce, but just 1 percent of those in computing. On a personal level, she notes, tech can be transformative. “Tech is amazing because you have the opportunity to change your life and change the trajectory of your family and your people. More so than other industries. Way more. Way more.”

In the world of tech entrepreneurs, Brittany Greenfield also sees very few people like her. In 2022, companies founded by women took in 2.1 percent of venture capital dollars, according to Pitchbook, while women-founded companies with an AI focus represented 0.3 percent of AI investment, according to a report released last month by The Alan Turing Institute. These are numbers so vanishingly small that women can feel forced to refashion themselves as male entrepreneurs.

It isn’t uncommon for venture capitalists to advise female founders to “triple your revenue projections” because that’s what a man would do. Greenfield remembers a venture capitalist telling her, “‘They’ve done research. Women are going to tell it like it is, and unfortunately the VC models have been trained to assume that you’re lying by [three times]. You need to get over the fact that you’ve put out a real projection — just take whatever number you’ve come up with and multiply it by 3x.’”

Bringing more women into tech is a work in progress. By 2021, the percentage of women earning undergraduate computer science degrees had grown to 21 percent. The number of underrepresented women — Black, Latina, and Native American — getting computing degrees has nearly doubled over the past five years, according to Pivotal. But tech is absorbing so many graduates that their percentage of the workforce is actually slipping. When it comes to senior leadership, men still occupy about 90 percent of the roles. It’s a reality Greenfield observes at conferences.

As artificial intelligence begins to remake computing, women have a new window of opportunity — though it won’t be open for long. El Kaliouby says that AI’s ubiquity could “allow more women to leverage these technologies to solve problems they care about.” But if women don’t become AI leaders, “you’re solving only a small subset of problems in the world. You’ll leave a set of problems behind.”

Even more worrying, she says, is that if women aren’t founders in the next wave of companies, they won’t become the next generation of deep-pocketed investors, who can then invest in other women founders. “Technology, for me, was a vehicle for social and economic mobility, and I’m very grateful for that,” says el Kaliouby, who notes that she grew up in a solidly middle-class Egyptian family before traveling to England and then the United States to further her studies. “And I feel that if women are not part of this AI revolution, it will have dire economic and financial implications.”

Kara Miller writes The Big Idea column for the Globe. Follow her @karaemiller. Send comments to magazine@globe.com.

Follow Kara Miller @karaemiller.