Session Submission Summary
Share...

Direct link:

Stochastic Parrot or Creative Partner?: Artificial Intelligence (AI) in Slavic Literary Studies & Translation

Fri, November 21, 3:30 to 5:15pm EST (3:30 to 5:15pm EST), -

Session Submission Type: Panel

Brief Description

The past few years have seen intensified debates over whether AI will ever produce good art. Bender et al. (2021) famously describe large language models (LLMs) as “stochastic parrots,” arguing that these models mimic form without comprehending meaning, necessitating a critical evaluation of any claims made about their creative capabilities. Michele Elam (2023) suggests that LLMs not only fail to grasp meaning but actively “render meaning senseless,” complicating our ability to comprehend and articulate our own reality.

However, Elam goes on to argue that, when engaged critically, AI might “productively challenge the arts” by raising new questions about authorship, creativity, and the arbitration of taste. Among the cautious optimists, Katherine Elkins (2024) suggests that advancements in LLMs offer new insights into how language—including literary language—encodes knowledge. Evan Donahue (2023) proposes that literary scholars have vital expertise to contribute to AI research, which is increasingly concerned with questions of authorship, genre, and how form mediates perceptions of reality.

This panel is a response to urgent calls for humanities scholars to take part in these conversations (Kirschenbaum and Raley, 2024; Ledesma, 2024). The panel aims to explore the repercussions of these debates on our discipline and to consider the role Slavic Studies scholars might play in the growing dialogue between humanists and AI researchers. Specifically, it seeks to ascertain what unique perspectives from Slavic literary theory and scholarship might be brought to bear on these broader debates and asks whether recent developments in LLMs might inform our understanding of Russophone literary creation.

Sub Unit

Chair

Papers

Discussant