Just over a year ago we envisioned a conference that could bring together an interdisciplinary group of people to talk about what we believe are critically important topics: responsible and society-centered AI. This inaugural event spurred the creation of a community of people inside and outside of Duke who are enthusiastic about connecting and collaborating on society-centered AI.
This year, we wanted to expand the reach of the conference. We asked ourselves:
At this year’s Conference on Society-Centered AI, we tackled these questions. Over the 3-day conference (and 24 hour hackathon the weekend before!), we convened 950 registrants from 50 universities and 80 companies. We hosted leading researchers, educators, and community members exploring different facets of society-centered AI.
We incubated new explorations in society-centered AI: A week before the conference, we hosted the Society-Centered AI Hackathon. Over 150 students registered and 19 teams ended up completing the 24-hour challenge: how to evaluate society-centered AI. The diversity of approaches was outstanding! Check out the amazing winning projects below:
A huge thank you to our wonderful judges, George Dohanich (Google Cloud), Funda Gunes (Walu AI), and Hazem Mahmoud (Patrick J McGovern Foundation)!
We highlighted the incredible scholars working on these problems at Duke and beyond:
Keynote Talks from leading scholars and innovators, discussing what we are NOT evaluating with traditional AI evaluations (and what we should be doing) (Arvind Narayanan), how interactions with AI/robots are shaping human behaviors (Nicholas Christakis), and how we are all creating the story of AI (Reggie Townsend).
The fan favorite “New faculty talks” on cutting edge research - disciplines represented included public policy, engineering, computer science, ecology, sociology, english, and philosophy! Featuring Rich Eva, Anne L. Washington, Daniel Scott Smith, Wenhao Jiang, Shuyan Zhou, Tong Qiu, Alessio Brini, Richard Jean So, and Brandon Fain.
Research abstract spotlights from Duke and neighboring institutions on topics ranging from the ethics of artificial companionship to interpretable approaches for visualizing high dimensionality data: Ateba Whitaker (Virginia Tech), Gaurav Rajesh Parikh (Duke University), Margaret Traeger (University of Notre Dame), Mohammad Hossein Jarrahi (UNC Chapel Hill), Delaney McLaughlin (Duke University), Jiechen Li (Duke University), Jae Yeon Kim (UNC Chapel Hill), and Yunxi Kong (Duke University).
We made the conference even more interdisciplinary through:
Incredible panels featuring panelists who were thoughtful, bold, and courageous. They discussed topics that people often shy away from: AI intimacy and companionship, the ambiguity of AI use in the music industry, and why we aren't doing more to support community engagement with AI. The panelists included: Walter Sinnott-Armstrong, Jana Schaich Borg, Kaamna Bhojwani, Timothy Strauman; David Hoffman, Tift Merritt, Ted Kalo, Christopher Buccafusco; Sanyin Siang, Siobahn Grady, Henry C. McKoy, Jr., and Natalia Summerville.
Something incredibly important to us is the inclusion of scholars at all levels, so we were thrilled to include:
Research talks from our top scoring student scholar abstracts followed by a panel led by Neil Gaikwad. The brilliance and thoughtfulness of these young researchers gives us a lot of hope for the future: Tiffany Degbotse, Rose Ansari, Shaozhe Cheng, Breanna Nguyen, Jiaxun "Jessie" Cao, and Frederick Coleman.
A student-led "reverse panel" where students Lindsay Gross, Peter Banyas, and Neha Shukla asked Open AI Chief Economist and Duke professor Aaron "Ronnie" Chatterji thoughtful questions on how AI is being used across the world, AI and the economy, and the future of work.
We built community around society-centered AI:
Across the three days, we were overwhelmed by the thoughtful questions and reflections posed by the audience. This extended into our daily poster & networking sessions, where we could all grab a mocktail and share ideas with each other. This year, we were thrilled to also feature art exhibits at our poster sessions. The buzz in the poster sessions was palpable, and people lingered long after the end of the sessions, connecting and sharing ideas.
What’s Next?
For information on our upcoming events, please subscribe to the SCAI newsletter. We look forward to seeing you again next year!
This post would not be complete without a huge thank you to all of the people who made this possible: our wonderful speakers, hackathon judges, coordinators (Rachel, Amanda, and Blake), “DJ” (Kevin), videographers (Big Toe Media), photographer (Ryan) and support from Fuqua’s event staff (Wendy).
Our co-sponsors for the event were generous with their support: sharing resources and time with us. This would not have been possible without the support of the Fuqua School of Business, Duke Science and Society, the Duke Center for Computational Thinking, the Duke M.Eng in Artificial Intelligence, the Coach K Center on Leadership and Ethics, Duke CASE, and the Duke M.Eng in Design & Technology Innovation.
Finally, we had over 50 (!) of amazing student volunteers that supported the event, bringing their time and energy to the event. We are so appreciative of your help in building this community.