Main content
AI.Humanity initiative announces spring 2024 seminar series
Media Contact
Ashlee Gardner
Julia Wrobel, Phillip Wolff, Peter Hitchcock, Prasanna Parasurama

Left to right: Julia Wrobel, Phillip Wolff, Peter Hitchcock, Prasanna Parasurama

The Emory community is invited to delve into the ethical, societal and scientific implications of AI and machine learning technology through the spring 2024 AI.Humanity Seminar Series. Sponsored by the Office of the Provost and the Center for AI Learning, the series will highlight new research and interdisciplinary perspectives via engaging and accessible TED-style talks delivered by AI.Humanity faculty.

The seminars are open to the general public. Attendees will have the chance to mingle and enjoy light refreshments before and after each seminar.

Explore the full schedule of talks below.

Julia Wrobel: “Using Digital Biomarkers to Detect Recent Cannabis Use and Cannabis Impaired Driving”

Wednesday, March 6, 2 – 3:30 p.m.
Convocation Hall, Room 208

Marijuana is now legal for recreational or medical use in 41 states. Accordingly, there is a need for objective and validated measures of acute cannabis impairment that may be applied in public safety and occupational settings, such as post-crash or accident investigations. This talk discusses statistical and AI models for detecting recent cannabis use, as well as potential ethical implications.

Julia Wrobel is an assistant professor in the Department of Biostatistics and Bioinformatics in the Rollins School of Public Health. Previously, she held a faculty appointment in the Department of Biostatistics and Informatics at the Colorado School of Public Health. Wrobel obtained a BA in chemistry from Swarthmore College in 2010 and a PhD in biostatistics from Columbia University in 2019. Her research interests include analysis of single-cell imaging data, functional data analysis, data registration and normalization, cancer imaging and analysis of data from continuously streaming sensor devices.


Phillip Wolff: “Leveraging AI to Predict Psychosis”

Wednesday, March 20, 6 – 7:30 p.m.
Convocation Hall, Room 208 

The domain of speech analysis for the early identification of schizophrenia is rapidly advancing. In this presentation, Wolff will showcase cutting-edge research demonstrating the ability of Large Language Models (LLMs) to predict the early signs of schizophrenia, drawing on data from the international Accelerating Medicines Partnership® in Schizophrenia (AMP® SCZ) initiative. This new approach to the analysis of language significantly outperforms conventional methods, signifying a breakthrough in the field of behavioral phenotyping. Wolff will explore how the linguistic analysis enabled by AI technologies can deepen our understanding of human cognitive functions and help unravel the intricacies of how information is processed within LLMs. This exploration will shed light on the intriguing similarities between human mental processes and the operational principles of generative AI models.

Phillip Wolff is a professor of psychology in the Emory College of Arts and Sciences. He is affiliated with the Cognitive and Computational Science and Clinical Science concentrations. Additionally, he is core faculty in the Program in Linguistics and associated faculty in the Department of Computer Science. Wolff's recent research focuses on using natural language data mining to extract biomarkers related to psychosis, neurodegenerative diseases and fundamental cognitive processes. His interdisciplinary approach combines insights from psychology, computer science and linguistics. He earned his PhD from Northwestern University.


Peter Hitchcock: “Can Learning Principles from AI Transform Psychotherapy?”

Wednesday, April 3, 6 – 7:30 p.m.
Convocation Hall, Room 208

CBT, or cognitive-behavioral therapy, is a widely used psychotherapy, with roots in decades-old behavioral research. While effective for many, others experience slow progress or treatment failure, leading to long waitlists and high dropout rates. Our era of advanced AI offers an unprecedented opportunity to optimize CBT. Reinforcement learning (RL), a key area of AI, shares roots with CBT, yet the remarkable advances from this interdisciplinary RL have yet to be integrated into CBT. Establishing a translational pipeline from RL to CBT is a primary aim of Hitchcock’s research ( In this talk, he will discuss this goal and its distinction from the dominant focus in the field of computational psychiatry and describe a translational case study concerning how cognitive actions are learned, and the implications for psychotherapy for rumination and worry.

Peter Hitchcock is an assistant professor of psychology in the Emory College of Arts and Sciences. He is dually trained as a clinical psychologist with a focus on behavioral treatment and as a computational psychiatry researcher with a specialty in applying formal learning and choice models to depression and anxiety. He received his PhD in clinical psychology from Drexel University, completed clinical internship at the VA San Diego in 2019 and was a post-doc with Michael Frank at Brown University from 2019-2023.


Prasanna Parasurama: “Foundational Models and Risks of Correlated Failures: The Case of Algorithmic Hiring”

Wednesday, April 17, 2 – 3:30 p.m.
Convocation Hall, Room 208

Demand-side hiring choices on LinkedIn are influenced by the platform’s recommendation algorithm. The recommendation algorithm, by design, takes into account not only the probability that an employer will contact a candidate but also the probability that a candidate will respond positively to that contact. Prior scholarship has shown how supply-side choices are influenced by anticipation of discrimination in the hiring process. If minority workers are less likely to respond positively to a LinkedIn contact anticipating discrimination, then the platform will be less likely to recommend them to the employer.

This talk will examine whether this influence of supply- side choices on demand-side choices on LinkedIn contributes to occupational segregation. 

Prasanna Parasurama is an assistant professor in the Department of Information Systems & Operations Management at Emory University’s Goizueta Business School. He received his PhD from NYU Stern School of Business and his undergraduate degree from the University of Michigan. Prasanna’s research areas include algorithmic hiring, algorithmic bias and fairness and human-AI interaction. His research leverages a wide array of quantitative methods including econometrics, machine learning and natural language processing.

Recent News