AI Poetry and Human Language (Joint Discussion) (via Zoom)

Abstract: Traditionally, people's ability to use language is attributed to abstract knowledge of how to manipulate symbols to efficiently convey information. However, widely different sources of evidence show this assumption to be misguided. In this talk, we focus on one of them: recent advances in deep language models, specifically the extraordinary capacities of OpenAI's GPT3. We will present in detail a current project exploring its creative capabilities in a particular realm, poetry, and discuss what this means for the assumptions of traditional theories of language.

Bio: Morten H. Christiansen is the William R. Kenan, Jr. Professor of Psychology at Cornell University, Professor in Cognitive Science of Language at the School of Communication and Culture as well as the Interacting Minds Centre at Aarhus University, Denmark, and a Senior Scientist at the Haskins Labs. He was elected as a foreign member of the Royal Danish Academy of Sciences and Letters, as well as elected Fellow of the Association for Psychological Science and of the Cognitive Science Society. Christiansen is the author of over 225 scientific papers and has edited four books and authored two monographs. His newest book aimed at a general audience, The Language Game: How Improvisation Created Language and Changed the World, outlines a radical new perspective on how language works. 

Bio: Pablo Contreras Kallens is an MA in Philosophy and a PhD Candidate at the Psychology Department in Cornell University. He's part of Dr. Christiansen's Cognitive Science of Language Lab. His main focus is on computational explorations of theories of language learning, with a particular emphasis on statistical learning.