Talk Title: 

Future Shock: Grappling with the Generative AI Revolution

Talk Description: 

The title of this presentation is taken from a forthcoming special issue of the Harvard Data Science Review. More than half a century ago, sociologist Alvin Toffler coined the term “Future Shock” to describe the dizzying disorientation brought on by the premature arrival of the future.

While the debate continues whether the advent of generative AI constitutes a future shock, its emergence presents a fascinating array of grand challenges and dilemmas.

Is generative AI intelligent to some degree? How can we ensure confidentiality protection in AI? Is it possible to protect individual privacy, and what does “individual privacy” even mean in the context of generative AI? This talk invites the audience to contemplate these issues, emphasizing the necessity of invoking multiple perspectives, particularly philosophical, legal, and statistical. While the findings may not be shocking, they certainly push us beyond our comfort zones.

Speaker Bio:

Xiao-Li Meng, the esteemed Whipple V. N. Jones Professor of Statistics and the innovative Founding Editor-in-Chief of Harvard Data Science Review, is renowned for his remarkable achievements in research, pedagogical innovation, administrative vision, and engaging communication style as both a speaker and writer. 

Honored as the best statistician under 40 by COPSS (Committee of Presidents of Statistical Societies) in 2001, Meng has received numerous awards for his prolific contributions, spanning over 150 publications across theoretical, methodological, pedagogical, and professional development realms. His captivating column, "The XL-Files," in the IMS Bulletin reflects his thought-provoking and entertaining writing. Meng's expertise covers theoretical statistical foundations, statistical methods and computation, and practical applications in natural, social, medical sciences, and engineering. 

Graduating with a BS in mathematics from Fudan University in 1982 and earning his PhD in statistics from Harvard in 1990, Meng has had a distinguished career, including serving on the faculty of the University of Chicago from 1991 to 2001. Upon returning to Harvard, he held prominent positions such as Chair of the Department of Statistics (2004-2012) and Dean of the Graduate School of Arts and Sciences (2012-2017).

Talk Title:

AI in Medicine

Talk Description:

Dr. Ohno-Machado will present an introductory seminar on the potential consequences of the increased use of predictive models in healthcare. With the increased sharing of clinical data, many issues related to privacy, validation, and representativeness of AI models have surfaced and must be addressed. Biomedical informatics and data science have a major role in ensuring that models first, do no harm and second, provide measurable benefits to individuals, groups, and society in general. 

Speaker Bio:

Dr. Lucila Ohno-Machado, MD, PhD, MBA, holds the positions of Deputy Dean for Biomedical Informatics and Chair of Biomedical Informatics and Data Science at Yale University. 

In her role, she oversees strategic initiatives in biomedical informatics research within the academic health system. The Biomedical Informatics and Data Science department at Yale is a hub for collaborative efforts in the biomedical field, integrating informatics into clinical practices and pioneering innovative approaches to analyze vast datasets. Dr. Ohno-Machado's distinguished career includes leadership roles at the University of California San Diego (UCSD), where she played key roles in informatics and technology. 

Her academic journey includes degrees from the University of São Paulo, Brazil, Escola de Administração de São Paulo, Fundação Getúlio Vargas, Brazil, and Stanford University. Recognized as an elected member of prestigious organizations, she has received accolades, including the American Medical Informatics Association leadership award. 

Driven by a fascination with the intersection of life science and computer science, her research focuses on predictive models and data sharing, including neural network models for survival analysis and innovative algorithms for federated computation.