By SHANNON O. WELLS
In his presentation to Faculty Assembly on updates regarding Pitt Information Technology (IT), Mark Henderson, vice chancellor and chief information officer, called the 2024-25 academic year “sort of a journey of transformation.”
“There are a lot of things that are moving us to that,” he said at the all-virtual May 7 meeting. He cited six challenges in particular:
-
Changing student expectations. “Every year, new students come here with different expectations around what they want out of their technology experience at Pitt.
-
Cyber risks and data privacy “are going through the roof.”
-
Political uncertainty coming out of Washington and beyond
-
Competition and enrollment pressures. “We’re looking at growing our classes year over year.”
-
Cost pressures, with cuts from the National Institutes of Health and National Science Foundation as two examples.
Changes in technologies “causing some skill gaps.” These are affecting some positions that “we relied very heavily on,” he said, including “positions and capabilities that we can, quite frankly, combine as all-inclusive services from some of our vendor partners. We’re looking at that, and then where we have needs, we have to come up with plans to enhance our team members’ skills so they’re able to help us in some of these emerging areas.”
Henderson said they are “working very hard to evolve from a traditional IT organization to one that can become more nimble, more responsive, and ideally, in time, anticipatory of faculty needs and student needs.”
This requires building a “commonality of process, tools and platforms” including simplifying and utilizing automation wherever possible, “because there’s still an awful lot of things that we do manually that need to be automated.”
Pitt IT also is working to transform itself in areas of utilizing data systems and solutions.
“Instead of a traditional view of data warehouses, data lakes, what we started thinking about is a universe of data, because universes are ever expanding, and our search for data is never really quenched, given the different uses and different data sets that are available to us.”
Rather than creating a larger research enterprise organization, the transformation process involves making the center “much smaller, while pushing those who are part of Pitt IT out to live and breathe every day in and among the faculty, students and staff that we serve,” he said. “We’ve done a lot in that regard.
“And to really provide AI tools and technologies that enhance our community. We’re going to do that, always with appropriate use and guidelines, because things are happening too fast for us to look at establishing some hard, fast policies in this space.”
Homegrown intelligence
Considering benefits to Pitt faculty is a key part of the transformations Henderson shared.
“We want to use our transformation to enable new teaching technologies and tools and approaches … to include assistive technologies as well for those who might have challenges,” he said. “We want to, over time working with faculty members and our friends at the Center for Teaching and Learning, look at how do we create personalized learning experiences. … Digital tools offer opportunities for that.
With the launch of Pitt EDGE, IT is examining how it leverages tools and technologies to “enhance the delivery of content … campuswide. How do we continue to improve and enable the research that you all engage in through the use of tools and technologies — AI and machine learning not being the least of that,” Henderson said.
Some recent initiatives include an AI enablement team focused on, “How do we understand some of the emerging needs, and how do we evaluate some of the tools and technologies that are out there in the AI space. And how do we best bring them to bear here at the University,” he said.
Another area of “great interest,” involves how to ensure AI literacy for IT clients.
“How do we make sure that people understand what’s good about them, what might not be so good about them, and what works for the University, and what may or may not work for individual-use cases,” Henderson explained, noting a deal signed last fall with NVIDIA, Carnegie Mellon University and the state of Pennsylvania around creating NVIDIA’s first AI tech community lab in the Americas.
“That’s pretty exciting,” he said.
A lot of IT work is being undertaken in concert with Rob Rutenbar, senior vice chancellor for research, on various projects and opportunities. Pitt IT also has spent a lot of time working with Pitt AI Scholars and Teachers Association (PASTA), “and that’s been very interesting.”
The creation of Pitt GPT makes it possible for Pitt IT to not use its internet protocol (IP) to “turbo charge businesses out there,” Henderson explained, noting that using Microsoft Copilot or other free software “feeds their large language models and becomes part of their universe of data.
“So we thought the best thing for our environment would be to create our own GPT that exists within a Pitt University domain,” he said, “and is not accessible by other practitioners.
“We’ve been working hard with about 700 folks from across the University to really exercise this technology. We’re looking to make it production-worthy over the summertime, so that we’re good to go at the start of the new school year.”
Looking ahead
Henderson encouraged Faculty Assembly members to review Pitt IT’s 10-year planning horizon/IT enabling plan.
“We chose a 10-year horizon because it gives us really a chance to think broadly about what the future may hold from a technology standpoint,” providing a “strategic roadmap” for using technology at Pitt.
The feedback the plan has received shows that IT needs to be light on its feet, “because things are changing, not just the technology landscape, but the needs of our faculty, the needs of our students … And these times are very uncertain, so we need to be able to react where we can anticipate some of these changing needs.”
Responding to a question from Melanie Scott, vice chair for research in the Department of Surgery, about Pitt IT’s involvement with Google, which has a company presence in Pittsburgh, and its Notebook learning language model, Henderson affirmed the partnership.
“And in fact, we will be making available a number of things that we have as part of our enterprise license agreement with Google in their GPT space,” he said. “We’ll be turning some things on over” this summer.
Scott also asked about why a 10-year planning horizon was chosen for “something as fast-moving as AI?”
“The way we started looking at it was kind of a five and five,” Henderson explained. “The first five it kind of aligns with many elements of the Plan for Pitt. And in fact, we don’t have a separate IT strategic plan. We have an IT enabling plan.
Based on conversations with Rutenbar, Pitt IT decided to “look even beyond five years so that we … don’t get comfortable and complacent in the here and now … particularly through AI.”
Shannon O. Wells is a writer for the University Times. Reach him at shannonw@pitt.edu.
Have a story idea or news to share? Share it with the University Times.