Here’s why AI needs a more diverse workforce

Eleven year-old Gia Mar Ramos was furious. Her dad had signed her up for a robotics class at the Johns Hopkins Center for Talented Youth camp without telling her. When she arrived, there were 11 boys and only 1 other girl in that class. All the boys had previous experience with programming. The two girls decided to team up and try to beat the boys – which they did – in every single challenge. That experience motivated Gia to teach other girls about opportunities in science, technology, engineering and maths (STEM).

Diversity in tech: still a long road ahead

Gia’s initial sense of solitude is not unique; diversity in the technology industry has a long way to go. Despite talk of greater inclusion, women’s representation in tech-related jobs has declined 32% since 1990. In a 2014-2016 study, women held only 25% of computer jobs, roles defined as computer scientists, systems analysts, software developers and programmers. While there is still insufficient data on how the COVID-19 pandemic will impact these figures over the long-term, recent reports indicate that women are losing ground in the workforce. The percentage of black and Latinx employees at major tech companies is also especially low, about 1-3% of the tech workforce. Diversity is also lacking among entrepreneurs and the venture capitalists who invest in them.

Why diversity in AI is important: future implications

Artificial Intelligence (AI) systems are designed to assist and amplify human behaviour, such as problem-solving, learning and planning. The tech industry is increasingly adopting AI to enhance performance for many companies – from data sorting to predictive analytics. In fact, overall investing in AI is expected to grow to more than $190 billion globally by 2025.

Fueling this aggressive growth are human technologists; people needed to build these systems, and perfect the algorithms informing the decisions that AI is making. Yet these technologists are not nearly diverse enough to reflect the concerns of the populations they impact. Thus, as companies and governments look to the future of AI and jobs, they will need to ensure that those who are developing AI technologies don’t perpetuate or accelerate algorithmic bias. By addressing diversity gaps today, tech leaders can mitigate bias in the systems built for tomorrow. Doing so requires creative thinking and unconventional approaches.

Building a pipeline of diverse AI talent

One such effort took place this summer. The authors of this piece came together with our teams to chip away at inequity in the AI space with an educational initiative for high school and college students.

Dr. Rana el Kaliouby is the Co-Founder and CEO of the Emotion AI company Affectiva. In previous years, her team would select a small number of young people to participate in a coveted in-person AI summer internship program at their HQ in Boston, MA. When the COVID-19 pandemic hit and social distancing took hold, the programme was slated to be canceled.

El Kaliouby met Ronit Avni, a fellow female tech founder and CEO, through the Young Global Leaders network of the World Economic Forum. We share a passion for applying technology to social problems at scale and began to brainstorm. Avni had founded Localized, a talent tech platform that connects university students and recent graduates with industry experts to guide them, and employers to hire them. We both subscribe to the adage, “you have to see it to be it”. We know how important it is for young people to access “proximate role models”; diverse role models that students from any background could look at and credibly imagine, “if they can do it, maybe I can, too”.

Avni suggested that el Kaliouby move Affectiva’s programme online and scale it using her platform. Thus, a five-week virtual AI education bootcamp called EMPath was born, which brought together 53 students that were intentionally diverse, with regard to cognitive styles, gender, ethnicity and culture, to increase their collective productivity and creativity. Some had no skills in computer science or machine learning, others attended MIT. The collaboration between Affectiva and Localized focused on accessibility, cooperation, education and mentoring with the goal of developing future AI leaders who would build ethical, accessible AI solutions to societal problems.

The students got hands-on experience with machine learning, computer vision, data analytics and visualization; they had to collaborate across time zones, all working virtually. After a couple of weeks of self-paced learning, the programme culminated in a “Make-a-thon”, where students applied what they had learned to develop an application of Emotion AI that covered areas such as the future of the transportation experience, reimagining virtual events, and an emotion-enabled Internet of Things (IoT). The winning team proposed an affordable, responsive mirror that combats social isolation in populations such as the elderly. The second-place winner used Emotion AI to provide feedback on engagement to virtual speakers, event organizers and audience members.

Gia Ramos said upon completion of the programme: “I am passionate about AI because it includes a lot of math, coding, and research – all the things that I like to do. This programme gave me first-hand insights and experience. I am looking forward to new challenges that building a career in AI will bring me.”

Moving forward: building a future we can all see ourselves in

One of the external judges, Keith Foster, Partner at Trend Capital, commented after the Make-a-thon pitch contest: “Overall, I was just blown away by the quality. I love that they surrounded themselves with people that believe in the problems that they were solving, because that allows for multiple perspectives... I’m incredibly humbled at the thinking and the diversity. And to be a part of that and watch the future; to hear them share their thoughts – it was a fantastic way to wrap up my week.”

As students like Gia go back to school, we are already in talks about how to make the programme bigger and better for what we hope will soon be a post-COVID future. A woman-led AI company is not what people imagine when thinking of tech leadership. Working together, we are committed to driving meaningful change by purposefully changing the stereotypes, encouraging fresh dialogue and building a future we can all see ourselves in.