AI4ALL: Diversifying the Future of Artificial Intelligence

When Stanford undergrad Ananya Karthik was a high school freshman, she was curious about technology, but didn’t know much about AI before she attended the 2016 Stanford AI4ALL summer program. Six months later, along with two AI4ALL classmates, she co-founded CreAIte, a neural art program targeting students from groups underrepresented in tech fields. Since then, CreAIte has introduced more than 500 girls around the country to coding basics, interdisciplinary technology, and peers who share their interests.

Harvard computer science undergrad Catherine Yeo attended Stanford AI4ALL’s inaugural program in 2015 as a high school sophomore. She went on to co-found PixelHacks, a hackathon that each year introduces hundreds of girls to tech and AI. She’s also the founder of Fair Bytes, a blog focused on fairness and ethics in artificial intelligence.

To Fei-Fei Li, Sequoia Professor in Computer Science and co-director of the Stanford Institute for Human-Centered Artificial Intelligence (HAI), Karthik and Yeo are the much-needed new faces of AI.

“Where AI is going depends on the humans behind it, the humans developing and deploying this technology,” Li says. “If we want the value of this technology to reflect who we are, we want the creators of this technology to reflect who we are.”

A Diversity Crisis

AI’s diversity — or lack of it — has long been a concern for Li. In 2015, while serving as director of Stanford Artificial Intelligence Laboratory (SAIL), she collaborated with PhD student Olga Russakovsky — now an assistant professor of computer science at Princeton University — to create a first-of-its-kind summer outreach program for high school girls to learn about AI and its potential to improve human life. Along with Dr. Rick Sommer, executive director of Stanford Pre-Collegiate Studies, the trio founded Stanford-based SAILORS. By 2017, with the help of tech CEO and investor Rab Govil, the program became a national nonprofit and was renamed AI4ALL.

Today, AI4ALL operates summer residential programs at 16 universities around the country, each one hosting a cohort of 20 to 30 high-potential young leaders drawn from groups underrepresented in AI, including women, students from racial/ethnic groups underrepresented in the field, lower-income students, and those identifying as LGBTQIA and non-binary. Experience in computer science is not required.

“We want to bring together and highlight new voices who will lead and shape AI,” says AI4ALL’s University Programs manager Jonathan Garcia. “We need to foster a community of diverse leaders in the field, because we understand that artificial intelligence impacts our daily lives — everything from who gets a job, to who gets a loan, to who can enter the country. These tools are made by people, and we’re not including some of the most underrepresented and marginalized people in the making of these tools and in the creation of this technology.”

Glimpsing the Possibilities

Stanford AI4ALL is led by SAIL, with collaboration from AI4ALL, Stanford HAI, and Stanford Pre-Collegiate Studies. Students spend three weeks on campus immersed in lectures, hands-on workshops, group projects, and field trips designed to showcase the wide-ranging potential of AI technology.

“This is opening a door to a world with boundless possibilities,” Li says. “They learn the foundational ideas of AI as a technology. They see the people living in this world, the faculty, the students, the postdocs, the researchers, the industry leaders, the colleagues who work in various aspects of AI. They’re exposed to ideas on the social impact of AI; they engage in deep conversations about the ethics of facial recognition and the impact of self-driving cars on our cities and energy.”

COVID-19 concerns this year required Stanford AI4ALL to be held virtually, but students were still able to connect online for lectures, events, and small-group research projects, where they learned about AI’s potential to benefit humans, said program director Juan Carlos Niebles, a senior research scientist at SAIL.

“I love to see the progression of how they grow. They come in thinking AI is about ‘Terminators’ and bad robots, and they leave with a very enriched perspective of what AI is,” says Niebles. “The last day of the program they talk about the project they’ve worked on, and it’s always amazing to see them speak with such confidence on the technology, the algorithms, and the problem they’re solving. They really embrace it.”

Karthik, now a rising sophomore at Stanford who grew up loving dance, writing, and art, remembers being excited by the intersection of creative expression and tech innovation and was intrigued to learn that AI had serious implications for society.

“That idea propelled me to study tech ethics and policy on my own time in high school, and it continues to be a really big area of interest for me today,” Karthik says. “In college, I’ve been furthering my study of tech ethics and policy, which I’m looking to potentially pursue in my college career.”

Building a Network

To date, 134 high schoolers have graduated from Stanford AI4ALL, joining a total of about 500 nationwide. Many go on to join AI4ALL’s alumni group Changemakers, which provides lifelong access to resources including continuing education, college and career preparation, and, perhaps most important, peer and mentor support. Ongoing contact with her mentors has been particularly uplifting for students like Yeo, now a junior at Harvard majoring in computer science.

“It’s been amazing to watch everyone’s progress and journey since 2015,” says Yeo. “I’ve built a community of relationships that’s sustained for many years. As they continue to make huge steps forward in AI, that’s been massively inspiring to follow. The fact that I know them and have previously interacted with them helps me feel that I can be like them one day, pushing for progress and change in AI.”

AI4ALL still faces hurdles: Selecting the most promising young candidates — sometimes from among students who may not have been exposed to computing or programming before — can be challenging; funding is an ongoing concern; and COVID-19 restrictions currently limit some activities. Yet the program’s successes are already visible, says Stanford HAI’s Li. Program graduates like Karthik and Yeo continue to create programs and projects using AI to benefit their communities, and 85 percent of program graduates have declared or intend to declare computer science as a college major.

“Increasing diversity in any field is a long-haul challenge,” says Li. “What keeps me awake is ‘how do we move the needle.’ We hope to move it for every one of these students, but also collectively we want to move the needle for our country.”

“A technology with this impact, that is this profound, needs full human representation to help it realize its potential, but also to recognize where to put on the guard rails,” she says. “It takes all of us.”

Stanford HAI's mission is to advance AI research, education, policy and practice to improve the human condition. Learn more.