With great pleasure we welcome Dr. Zoltan Majdik from North Dakota State University as the third guest in this series at the University of Tübingen. As co-director of the Center for the Study of Digital Society, his research focuses on rhetoric and the question of how complex topics—such as climate change or medical (mis)information—are communicated through large volumes of text. His methodology relies on computer-assisted approaches to text analysis that draw on language models and other text-processing infrastructures based on deep learning and neural networks.
There are several ways to interact with Dr. Majdik during his stay in late October, one of this is his public lecture on 27th October.
27.10.2026 | 18:00 c.t. | Room 119, Brechtbau (Wilhelmstraße 50)
The lecture focuses on the rhetorical tradition and how it can be used to examine AI systems. In his lecture, Dr. Majdik highlights four ways in which this can be done: as an object, as an instrument, as a rhetorician, and as a site of intervention. Based on this, he argues that persuasion and AI systems have less to do with individual persuasive power and more to do with the norms of rhetorical culture itself.
Abstract
In the wake of OpenAI CEO Sam Altman's prediction that AI will be "capable of superhuman persuasion," language models have become a dominant means of rhetorical production, generating persuasive language at a scale and in contexts that rhetorical theory has not accounted for. Yet the empirical record is more equivocal than Altman’s prediction suggests: a meta-analysis finds no overall persuasive advantage of AI systems over human persuaders, while leaving most of the variance in persuasive effectiveness across studies unexplained.
The largest study to date locates that variance in post-training protocols and prompting strategies rather than in personalization or scale. These are rhetorical variables. Drawing on studies of metaphor generation, epistemic stance, and rhetorical circulation, this talk maps four ways the rhetorical tradition can study AI systems — as object, instrument, rhetor, and site of intervention — and argues that what is at stake is less individual persuasion than the norms of rhetorical culture itself.
