This post was originally produced for Forbes.
The University of Southern California has recently created the Center on Artificial Intelligence for Social Solutions or CAISS, specifically to develop uses of artificial intelligence–AI–for use cases of interest to social entrepreneurs.
The Center is a collaboration between the USC Viterbi School of Engineering and the USC Suzanne Dworak-Peck School of Social Work. Eric Rice, an Associate Professor from the School of Social Work, has been tapped to lead CAISS.
The potential for AI to be an effective tool for entrepreneurs with no background in technology is confirmed by Rice himself. When asked about his LinkedIn profile URL, he acknowledged not only that he doesn’t have one, but also said, “I’m a bit of a Luddite.“
The Center has gathered commitments of $3 million for its launch. The two partner schools put up much of that money with additional funding from private partners. CAISS has an annual operating budget of $500,000 and has only one full-time staff person. The Center itself is a 501(c)(3) nonprofit.
Leading the new Center is as close to entrepreneurship as Rice admits getting, though as an academic he is entirely focused on social good.
Rice explains that the Center is looking at two particular sets of social problems. The first set is the “Grand Challenges for Social Work” established by American Academy of Social Work and Social Welfare. These twelve challenges are focused on social issues facing the United State. The other set is the Sustainable Development Goals or SDGs established by the United Nations in 2015 as goals for poverty reduction by 2030, primarily focused on the developing world.
One of the first projects the Center has undertaken is developing an AI tool to identify peer leaders among Los Angeles County’s homeless youth. By better identifying peer leaders, they can spread information about AIDS infection and prevention more rapidly. Their initial study showed success in increasing the number of youth who were getting tested for HIV.
Alison Hurst, the founder and executive director of a Safe Place for Youth, admits she hasn’t given much thought to how AI can be used to serve the homeless youth her organization is charged with helping. “I just know that we need to use all the tools in the bag to understand the interventions needed as our social problems keep growing,” she says.
The Center is one of the first places that AI engineers and social scientists have come together. Rice points out that “AI lets you model the messiness of the real world so you can probabilistically figure out how to proceed.”
One of the partners in the youth project is called My Friend’s Place, which is led by Executive Director Heather Carmichael, LCSW. She notes the potential for both financial and social benefits from implementing AI tools. “My Friend’s Place is doing amazing work with 1,500 youth experiencing homelessness every year,” she says.
“With limited resources and the breadth and complexity of the young people’s needs, it is our obligation to pursue knowledge, support and interventions that will reach the greatest number of youth,” she continues. “Imagine, if AI can help us identify 1 of the 100 youth we serve daily as a potential peer leader, we can expedite an invitation to health education and peer leadership programming, and ultimately ‘produce’ peer leaders socially positioned for the greatest impact!”
Rice observes that one of the biggest challenges the Center faces is getting the engineers and the social scientists speaking the same language. “Their way of thinking is very mathematical. Our way of thinking isn’t,” he says. “We had problems because we weren’t speaking the same language.”
That challenge also is the key to AI’s promise, Rice says. ” They ask each other questions they’ve never heard before .” This creates opportunities for answers they’ve never given before.
Not all social problems appear to lend themselves to AI applications, Rice acknowledges. Mental illness is an example he sees where AI interventions may be a long way off.
He insists, however, that AI has great potential for solving social problems and points at the success with the youth program. AI helps social scientists see unlikely outcomes, in contrast to traditional statistical models that help us see how the average person sees average problems.
Barbara Grosz, Higgins Professor of Natural Sciences in Harvard’s Paulson School of Engineering and Applied Sciences, also collaborates on the youth project in LA County. She is optimistic about the collaboration’s promise to help solve social problems.
“AI has the potential to support significantly, and in diverse ways, the work of people who are addressing social problems,” Grosz says. “For it to succeed in doing so, though, requires the combined efforts of those with expertise in AI and people whose expertise in the social sciences and policy give them a deep understanding of these social problems, their roots, and the key characteristics of approaches that are likely to work.”
She also highlighted the contributions of Rice’s co-founder, Milind Tambe. She adds, “It needs sufficient support to enable their work.”
While the collaborations have yet to create consumer products, it is clear that the Center’s work could now begin to complement efforts of social entrepreneurs with sufficient backing from impact investors to develop new tools for addressing social problems.
On Thursday, February 2, 2017 at 4:00 Eastern, Rice will join me here for a live discussion about using AI for solving social problems. Tune in here (at the top of this article) then to watch the interview live. Post questions in the comments below or tweet questions before the interview to @devindthorpe.