As public opinion sways between the benefits and risks of AI’s rapid rise into everyday life worldwide, UCSB Library’s “AI in Action” speaker series returns this fall with insight from researchers on how the technology is reshaping modern computing and human cognition. 

Following presentations by computer scientist Arpit Gupta and vision scientist Miguel Eckstein about their innovative use of AI as a research tool, a moderated discussion will field questions from the audience and speak to broader industry trends, challenges and ethics.

Gupta's presentation, “Making the ‘Net’ Work in the Agentic Era,” will unpack how his team at UC Santa Barbara’s Computer Science Department uses AI to accelerate networking research, inform broadband policy and move toward networks that run themselves.

Agentic systems push in two directions at once, Gupta said. “They bring that goal closer, and they speed up the research itself by taking over grunt work that is repetitive and slow, freeing researchers to spend their attention on harder questions.”

“The thing I care about most is the gap between a research paper and something anyone can actually use,” he added. “A paper takes insight and some engineering. A tool that another research group, a network operator or a regulator can rely on takes the same insight plus years of engineering. Agents are best suited to compress that extra work. Our team has been leading this effort, using agentic systems to deliver easy-to-use research artifacts for the scientific communities as well as policymakers and advocacy groups.”  

In “Why We Look Where We Look: How AI Answered an Age-Old Human Mystery,” Eckstein will reveal how building human visual traits into an AI agent helped solve long-standing questions about automatic gaze patterns that humans perform without conscious thought.

“AI has provided new powerful methods to study rich perceptual experiences that we were unable to study before,” said Eckstein, a distinguished professor in the Department of Psychological and Brain Sciences. “We can now present a complex scene to a human and study how vision supports their ability to understand and describe the scene. This is not something I envisioned a few years back.”

“In the last few years,” Eckstein added, “our lab has also analyzed the inner workings of AI models to gain insight into how attention works and make predictions about new attention-related neuron types that had not been previously noted by neurophysiologists.” 

Free and open to the public, the event starts at 4 p.m. on Tuesday, Sept. 29 in the library’s first-floor Instruction & Training Room 1312. Advance registration is recommended.  

Launched in early 2026 by the library and the UCSB AI Community of Practice’s AI for Research special interest group, the series aims to foster open, interdisciplinary conversations to deepen communitywide understanding of the technology’s capacity to expand access to knowledge and inspire new forms of scholarship.

This story was first published by The Current.