Cultural AI Lab

Duke University

AI slop

We are developing a science of unwanted AI content. So-called AI slop now saturates our online ecosystems, degrading trust, functionality, and authenticity — while at the same time carrying real affordances, opening strange new forms of creative and communicative expression. We build tools and methods to measure slop on platforms such as Reddit and Substack, parse its effects on user experience, and devise governance strategies that respect both its harms and its possibilities.

  1. Why Slop Matters Cody Kommers et al. ACM AI Letters, 2026. Read paper
  2. The Social and Cultural Life of AI Slop In progress.
  3. What’s to Like about AI Slop? Creating It Changes How You Feel about It In progress.

AI and creativity

All stories are social, and so is creativity. Current systems optimize toward a single generic reader, which silently targets the aggregate taste of the largest market and reproduces what is already dominant. We build instruments that hold reader difference rather than averaging it away — simulated communities of readers, grounded in real ones, whose responses are surfaced back to the writer, so that AI storytelling answers to actual audiences rather than to no one.

  1. Evaluating AI Creative Writing through Interpretive Communities In progress.
  2. Readers in the Loop: Building Interpretive Communities for AI Storytelling In progress.

Downstream effects of LLMs on digital platforms

We study what happens when LLMs are deployed at scale on live platforms like Reddit and Substack, and what their increasingly obvious problems — homogenization, model collapse — do to communication and creativity in the wild. Who benefits from deployment at this scale? How is it changing culture? Are humans starting to sound more like bots? Does detection work, or does it mostly punish the wrong people?

  1. Playing for the Detector: How AI Suspicion Changes What Artists Make In progress.
  2. Model Fingerprints: Which LLMs Are Writing the Internet? In progress.