Fall AI Town Hall brings updates on AI tools, computing and education
Attendees of the AI Town Hall meet in the Grand Challenges Research Building.
Photo by Leslie Hawthorne-Klingler, U of A Office of Research and Partnerships
The Office of Responsible AI hosted its fall AI Town Hall on Oct. 14, sharing updates on the University of Arizona's AI roadmap and introducing the goals of the new Task Force on AI in Education convened by the Provost and Chief Academic Officer and the Chief AI and Data Science Officer.
Presenters shared updates on the university's computing power, AI tools and training, noting that AI is already part of campus life and that everyone now has an opportunity to help decide how it's used in teaching and learning.
The task force is gathering input from faculty, staff and students. That feedback will shape a statement of principles, with a draft expected in early December, to guide how instructors and units approach AI.
"The goal of the university and the provost is the success of every student, and that means having the skills that will last you for the next 30 to 50 years," said David Ebert, chief AI and data science officer and leader of ORAI.
A campuswide conversation
Jennifer Savary, associate professor of Marketing at Eller College of Management and associate director of AI for education and curriculum at the AI Leadership Collaborative, introduced the goals of the task force. Across higher education, she said, some universities are embracing AI while others are restricting it, in some cases banning devices from classrooms.
Convened by Provost Patricia Prelock and Chief AI and Data Science Officer David Ebert, the task force includes representatives from the faculty senate, the undergraduate and graduate student councils, staff council and other campus groups. Its guiding question is how to strengthen students' ability to think, create, question and work with others while preparing them for a world reshaped by AI.
Drawing on surveys and earlier roadmap work, the task force identified four priorities to start the conversation: centering humans, AI fluency, expanding what's possible and responsible stewardship. Savary said the list is a starting point, not a final answer.
"Our true north here at the University of Arizona is always going to be centering humans," she said. "Our goal is the people."
After the draft principles circulate for campus comment, the task force plans to finalize them around February and develop an action plan by the end of the spring semester.
The tools and computing behind the work
Investment in the infrastructure that supports AI research is growing. Research computing is installing Lynx, the university's newest high-performance computing cluster, said Jeremy Frumkin, senior director for research technologies in University Information Technology Services. Lynx adds about 60 Nvidia H200 GPUs to the hundreds already available to researchers for AI and other demanding work.
Wolfgang Jentner, associate director of AI research applications, shares updates on the U of A GenAI tool, which has passed 6,600 users.
Photo by Leslie Hawthorne-Klingler, U of A Office of Research and Partnerships
Lynx joins Puma, the university's main cluster and Soteria, which supports research with HIPAA-protected health data. The U of A is also one of a handful of universities running a pilot cluster for controlled unclassified information, developed with the College of Engineering. As more funding agencies require federal security standards, the team is exploring how to turn that pilot into a full service. Researchers can find resources and support at the HPC documentation site.
The university is also expanding access to AI tools. The free U of A GenAI tool, which gives the campus access to several leading AI models, has passed 6,600 users since its rollout in February, said Wolfgang Jentner, associate director of AI research applications. Two new tools, the Arizona AI Assistant and the AI Discovery Lab, are in beta testing.
Those tools run on Amazon Web Services. Rob Cassetti, who leads AWS's work with state and local governments and higher education across 20 western states, said adoption at the U of A stands out in the region. "You've been incredibly thoughtful and purposeful, so it's not a surprise that the community has jumped on and is using the tooling that you've built," he said.
Through AI Innovation Pilots, faculty and staff can also work with research and software engineers to turn an idea into a prototype in eight weeks.
Building skills and confidence
Several units are helping the campus use AI with care. ORAI has hosted 11 AI fluency sessions this semester for about 656 participants, most of them staff. The University Center for Assessment, Teaching and Technology drew 468 registrants to its AI workshops last academic year.
Students can explore AI hands-on at the new AI+ Studio in the Weaver Science-Engineering Library, with library experts nearby.
"We're not pushing AI," said Tara Radniecki, associate dean of the University Libraries. "We want to give students room to explore AI. We want to give them room to ask hard questions."
In the College of Education, senior lecturer and AI fellow Austin Ross built online modules that help K-12 teachers use AI ethically. "We're not just using this because we can," he said.
Employees can find training in one place through the new AI Upskilling Hub.
The town hall led directly into AWS AI-Powered Research Days, two days of training led by the AWS team. Sessions shared lessons from AWS's work with universities nationwide and showed how the U of A's tools can be put to use.
