U students lead the conversation at inaugural AI symposium
“Google was processing 9.7 trillion tokens in May 2024. Now it’s basically 1.3 quadrillion tokens a month. That adoption curve is nonlinear no matter how you measure it.”
– William Barkley, vice president of product management, NVIDIA
UIT’s Digital Learning Technologies recently hosted a student-led symposium focused on artificial intelligence, bringing together University of Utah students, faculty, and staff, peers, and industry leaders for a full day of discussions, presentations, and networking.
“The purpose of this event is twofold: first, to give students an opportunity to share their perspectives and experiences; second, to provide space for nonstudents to consider what they hear in their decision-making,” said Jeb Dean, who is studying information systems. He co-chaired the event with Diya Mandot, a computer science major and Adobe creative consultant in DLT, Logan Bogesvang, a quantitative analysis of markets and organizations major, and Nicholas Pardon, an information systems major.
The inaugural Student AI Symposium, held November 21, 2025, in Marriott Library’s Gould Auditorium, kicked off with a keynote address by William Barkley, vice president of product management at NVIDIA; followed by a series of student presentations on how they use generative AI for their classes, research, and creative projects; a panel of university AI experts; and closing remarks from Sneha Kasera, Ph.D., associate dean for Academic Affairs and professor in the Kahlert School of Computing.
Barkley, an AI developer at Google for four years before moving to NVIDIA this past August, noted that unlike past waves of AI innovation, this one hasn’t slowed down.
Student lightning talks on AI projects:
- Munkh-Erdene Munkhbat: “AI X My Life”
- Reese Van Dyke: “Concerns on the ethics of AI”
- Campell Brown: “Can ChatGPT sentiment predict market returns?”
- John Bruns: “A new age of networking”
- Dominick Fighera: “DietWise diet tracker”
- Parker DeYoung: “Pantheon financial forecaster”
- Rishahb Saini and Diya Mandot: “AI isn’t our tool — it’s our reflection”
- Namase Kanehara: “AI use in the business world”
- Valen Cole: “Education models of the future”
- Heitor Augusto Mompean: “College applications using AI”
- Sadman Sunbeeb Islam: “Card Vault business card scanner”
- Kalista Leggit and Katherine Berensen: “A new AI crisis in mental health care”
- Benson Tracy: “Building AI systems that think in teams”
- Omid Zahr: “Custom DXF pricing system”
- Thomas Lu: “Solar/wind x ML technology”
“Almost five years later, we’re still moving at the speed of light,” he said.
Barkley explained key AI concepts like training, inference, and tokens. Tokens are raw bits of data that AI language models work with, with a million tokens equating to about 750,000 words. Referring to AI usage across Google’s products and services, he said, “Google was processing 9.7 trillion tokens in May 2024. Now it’s basically 1.3 quadrillion tokens a month. That adoption curve is nonlinear no matter how you measure it.”
A panel of leaders in the U’s AI community convened for a lively discussion on AI’s societal impact and how the university might prepare students for an AI-enabled workforce. Panelists included Manish Parashar, Ph.D., the U’s chief AI officer and executive director of the Scientific Computing and Imaging Institute (SCI); Penny Atkins, Ph.D., director for research and science at the SCI; Jim Agutter, associate professor in the College of Architecture + Planning; and Rebekah Cummings, Digital Matters director at Marriott Library.
Agutter captured the balance of the moment, saying, “You see a lot of optimism and opportunities for success, but to realize impactful results, we need to combine cautious optimism with taking control of AI to make sure we’re acting ethically and responsibly.”
“Faculty are asking, ‘Should we wall students off from tools or provide tools with guidance and clarity?’ We’re all sort of stuck in this weird moment of flux,” Cummings said. “We need to put funding into figuring out how we can teach most effectively with AI, to make sure we’re learning how to use [Microsoft] Copilot effectively, providing clarity on what’s a good use of AI, and being clear about what’s just a shortcut. It’s also important to teach students how to think without it. Day one on the job, they should be using AI efficiently.”
The symposium concluded with remarks from Kasera, who presented on “Building an AI campus,” in which he outlined a vision to make the U of U a top AI institution.
Kasera emphasized the growing demand for AI skills across myriad sectors and the need to embed AI education into every discipline.
Since hiring new faculty members specializing in AI is difficult, Kasera said the university has launched faculty training to integrate AI into research and teaching. Plans, he said, include discipline-specific AI emphases for undergraduates and tracks for graduate students, ensuring students can learn AI within their chosen fields.
“Our goal is to prepare our students for succeeding in their fields of choice, whatever that field might be,” he said.
Node 4
Our monthly newsletter includes news from UIT and other campus/ University of Utah Health IT organizations, features about UIT employees, IT governance news, and various announcements and updates.