Events

CDS Seminar: What Research Looks Like with AI Agents

Speaker: He He (NYU)

Location: 60 Fifth Avenue, Room 150

Date: Friday, March 13, 2026

Language models and AI agents are beginning to play a role in scientific workflows, raising questions about how research itself may change. Rather than focusing on the research frontier of co-scientists, this talk examines what is already possible today. I will present a case study of using AI agents to pursue a research question from initial exploration to experimental evaluation. The goal is to understand whether these systems can perform the kinds of routine intellectual and technical work that constitute much of everyday research. Based on this experience, I will share several practical takeaways for using AI agents as research collaborators. I will conclude with reflections on how these capabilities may affect the choice of research problems and the training of graduate students. I hope the talk will also serve as an internal discussion about how academic research may evolve as these tools become more capable.

Bio: He He is an Associate Professor of Computer Science and Data Science. She is broadly interested in natural language process and machine learning. Her recent research focuses on understanding large language models, improving their trustworthiness, and human-AI interaction. Prior to joining NYU, she obtained her PhD from University of Maryland, did a post-doc at Stanford, and spent one year at AWS working in dialogue platforms.