Amy Braverman, Ph.D.
Senior Research Scientist
Jet Propulsion Laboratory
California Institute of Technology
Technical Group Lead
Statistical Methods and Applications
Uncertainty Quantification and Statistical Analysis Group
Abstract. We all know that the landscape in which we do our work as statisticians is changing very fast. Two sea changes are upon us: one relates to data and one to how we analyze it. The one related to data is the “datafication” of the world around us. Data collection has become embedded in our everyday lives, from supermarket scanners that record what we buy to traffic sensors in the freeway, to clicks on web pages. These data are collected routinely, in highly decentralized environments, with no particular hypothesis in mind, and with uncertainties that depend implicitly on sensor technology. The second change is the exploding pervasiveness of AI. It offers new technical tools for statisticians to use but also creates a murky environment where prediction and understanding are easily confused. Perhaps more importantly though, AI assistants are changing the way we think and reason about problems while posing risks to that process.
Drawing on the recent National Academies report Frontiers of Statistics in Science and Engineering: 2035 and Beyond, this talk considers the interplay between AI and datafication, and their implications for the field of statistics. Two examples illustrate the challenges and aspirations of the new landscape. The first is uncertainty quantification for routinely collected remote sensing data, and the second is a look ahead at the next transformative development in datafication: the emergence of 6G edge networks.