Shazeda Ahmed (UCLA)

Abstract:

Dominant global geopolitical narratives about AI often rest on two tenuous assumptions: that the United States and China are locked in an arms race to determine who attains "artificial general intelligence" (AGI) first, and the only way AGI will not destroy all of humankind is if it is "aligned" with human values. While a growing body of interdisciplinary work addresses the myths and misrepresentations sustaining the former, there has been scant critical assessment of the "alignment problem" as it plays out in technical research in both the United States and China. This talk will present emerging sociotechnical research deconstructing theories and technical practices of AI value alignment research in the US and China, empirically surfacing the political challenges that emerge in this new subfield. 

Bio: 

Dr. Shazeda Ahmed is a Research Program Director at the University of California, Los Angeles. Shazeda holds a PhD from UC Berkeley’s School of Information. Previously, she has been a researcher at Upturn, the Mercator Institute for China Studies, the Citizen Lab, Stanford University’s Institute for Human-Centered Artificial Intelligence (HAI), the AI Now Institute, and Princeton University’s Center for Information Technology Policy. Her research has investigated how tech firms and the Chinese government are collaborating on the country’s social credit system, the political economy of emotion-recognition and courtroom AI technologies in China, and the epistemic culture of the emerging field of AI safety. Her work has been featured in outlets including the Financial Times, WIRED, The New York Times, and the BBC.