HCI Seminar - Ana Crisan - Do Humans Still Matter? Data Analysis and Visualization in the Era of Generative AI

Abstract:
Generative AI agents are increasingly able to produce code that can analyze and visualize a myriad of datasets. With this potent automating capability, it’s worth asking whether humans offer much more than an oversight role - if that at all. In this talk, I will explore the capabilities of GenAI systems for data analysis, highlighting both opportunities and core limitations of this technology. I especially seek to highlight the user experience and friction inherent in human-AI interaction when conducting data analysis and how this can impact the design of analysis tools. I conclude by drawing attention to the hard problems in data analysis and how these can be a yard stick to measure AI progress and human insight. 

Bio:
Anamaria (Ana) Crisan is an Assistant Professor in the Cheriton School of Computer Science at the University of Waterloo. Her interdisciplinary research integrates techniques and methodologies from artificial intelligence and human computer interaction to develop human-centered AI tools for data driven decision making.  Her research explores how visual and interactive interfaces can serve as a valuable medium for collaboration and information exchange between people and AI systems. Prior to joining the University of Waterloo, Ana held several industrial research appointments.  She was formerly a Lead Research Scientist at Tableau (2019 – 2024), a research scientist at the British Columbia Center for Disease Control (2013-2019), and Bioinformatician at Decipher Biosciences (2010-2013). Her award-winning research appears in top -tier venues of ACM (CHI, FAccT) and IEEE (Vis) in addition to highly ranked biomedical journals (Nature, Oxford Bioinformatics, PLOS).

This talk will also be streamed over Zoom: https://mit.zoom.us/j/96808776848.