Embracing the New Age of AI

At UBalt, we believe that AI is not a threat, but an opportunity.  Under the direction of the Center for Excellence in Learning, Teaching, and Technology (CELTT), the University is ensuring that the technology complements education rather than overshadowing it.

BY POORNIMA APTE
ILLUSTRATIONS BY MARIO WAGNER

For a long time, higher education has moved at a predictable pace with educators adopting new technology tools gradually. The advent of artificial intelligence (AI) feels much more disruptive, especially because its pace of evolution is so rapid.

Since news of the impressive capabilities of generative AI—ChatGPT is its most familiar variation—broke in November 2022, the technology has been making systematic inroads in higher education. A 2024 Turnitin survey of college students in the fall 2023 term found that more than half used generative AI in some manner.

The post-AI landscape is entirely new—and unprecedented. With respect to AI, “what I would offer is that The University of Baltimore, as other universities, is in the midst of what might be similar to the Cambrian explosion in prehistory, when you had so many life forms in such a small time and such accelerated evolution— you’re in the middle of something that dynamic, it’s unclear where you are relative to others,” says Dr. Alan Lyles, Henry A. Rosenberg Professor of Government, Business and Nonprofit Partnerships at the College of Public Affairs.

“Educators are having to make choices that are strategic, tactical, operational. They are committing resources and setting priorities perhaps for a decade ahead, but those decisions have immediate consequences,” Lyles says, “they’ve never done any of this.”

Julia Goffredi, instructional designer & emerging technology lead at the Bank of America Center for Excellence in Learning, Teaching and Technology (CELTT) agrees. The challenge, she says, is “how do you make meaning of something that’s so radically different from the way you’re used to presenting and disseminating information and knowledge.”

AI STRATEGY UNDER CELTT LEADERSHIP

Making meaning of this disruption is a challenge that CELTT is tackling head on, under the leadership of its director of teaching and learning excellence, Dr. Jessica Stansbury.

A doctorate in instructional technology had already equipped Stansbury with the tools for integrating new technology into classrooms. When she first test drove the public-facing version of ChatGPT, she was excited about its potential to help faculty. “My first thought was like, this is great for efficiency; faculty would be able to get rid of some of the tedious and mundane tasks they do,” Stansbury says.

But when not all educators seemed excited— they were even worried or skeptical about the technology’s ramifications—Stansbury knew CELTT had a more complex challenge on its hands. The problem: How to square faculty concerns with the University’s motto, “Knowledge that Works.”

Stansbury’s views on the acquisition of knowledge—“we are facilitators, not gatekeepers”— shapes CELTT’s approach to AI at UBalt.

“When you tie it back to [the University’s] vision, our goal is to make sure our students are prepared to navigate a modern workplace,” Goffredi says. And that modern workplace will need a knowledge of AI tools. “We’re setting up our students for success in society, so we have to recognize that AI is not going away and it’s going to be an AI world,” Stansbury says.

LAYING THE GROUNDWORK FOR AI INTEGRATION

Setting up students for success starts with the basic building blocks of AI integration, and UBalt is making a conscious—and intentional—effort to approach this technology proactively rather than reactively.

According to Stansbury, campus partners “will assess the immediate and emerging AI applications most likely to impact teaching, learning and research and explore the long-term needs of institutions, instructors, and scholars, as they navigate this environment.”

Among the first orders of business: joining the “Making AI Generative for Higher Education” initiative, launched by Ithaka S+R, a higher-education research firm. Along with 17 other member universities in its cohort, UBalt is in the midst of a four-phase, two-year research project that aims to establish the responsible use of generative AI on campus. Being part of the Ithaka project enables UBalt to observe how peers are faring and to establish guidelines and guardrails for AI use. While the culmination of the Ithaka project will be a concrete set of guidelines for the entire University to follow, CELTT has begun the work of developing guidelines for AI use now.

Understanding that not all members of the campus will adopt the technology at the same pace, CELTT has opted against a top-down AI policy. UBalt has instead taken a data-driven approach and has regularly polled its community and brought them along on the AI adoption journey. “Just as we wouldn’t administer medicine without clinical trials, AI policies should not be created without the expertise of educators and researchers, with all stakeholder voices in the conversation,” Stansbury has written.

To introduce faculty to AI, Stansbury arranged a “learn with me” session for instructors in January 2023, just a few months after ChatGPT’s public unveiling. After encouraging casual conversations about AI in early 2023, it was time for a more formal introduction.

CELTT recruited Amr Kadry, M.S. ’20, coordinator of tutoring and academic coaching services at UBalt’s Robert L. Bogomolny Library to develop courses for both faculty and students. Kadry applied for (and was awarded) an Elkins SoTL (Scholarship for Teaching and Learning) 2023-24 fellowship, granted by the University System of Marland’s William E. Kirwan Center for Academic Innovation to design and research the impact of these courses.

The asynchronous courses, called “From Chalkboards to Chatbots,” introduce the University community—including students and faculty—to the basics, which they can visit at their own pace. The gentle (yet structured) start was intentional. “We’re not fully pushing the tools and at the same time, we’re removing some of the fearmongering, the main purpose is to just expose everyone to AI and start the conversation,” Kadry says.

At the same time, says Goffredi, CELTT’s approach to AI literacy has been one that “dissipates the fear but keeps the curiosity. We didn’t want folks to just jump off the cliff. We need to make sure that everything’s tight and secure before you go bungee jumping,” Goffredi says.

Also key has been the focus on literacy skills and not any one platform, Kadry says. Such an approach is a reminder that the platforms might evolve—Bard today, Gemini tomorrow—but the fundamental technology behind AI is what educators and students will need to grasp. “We’re taking more of a mindset that it’s important for us to start integrating AI literacy into the UBalt educational experience,” Kadry says.

AI FOR EDUCATORS

University educators are also doing their part to integrate AI into the UBalt experience. They have participated in the CELTT courses and been active and vocal in shaping how the University approaches the technology.

“Being proficient in the use and the education of AI as a tool is the price of admission to being a professor today. Period. Hard stop,” Lyles says.

A few educators have also drafted policies for the technology’s use in their classrooms. “If there’s no AI policy then each educator is on their own and we’re unnecessarily introducing entropy,” Lyles says. To avoid the problem, he decided to craft an AI policy, which itself would be refined over the semesters. “I’m going to incorporate feedback as we go along but it gives learners a knowable target,” Lyles says, “it gives guardrails and also aligns with assignments and expectations.”

“ Being proficient in the use and the education of AI as a tool is the price of admission to being a professor today. Period. Hard stop.”
DR. ALAN LYLES

It’s not just about policy for student AI use. Having generative AI has forced educators to rethink assessments and how to measure what students are actually learning. “I was kind of happy when AI came along because it really shook up how we think about teaching,” Stansbury says. “If students can finish an assignment using just AI, then it’s probably not a very good assessment of learning.”

“The challenge for faculty is that for them to reimagine assignments, they have to have some knowledge of ChatGPT’s abilities and how it works,” says Dr. William Carter, associate professor of management in the Merrick School of Business and recipient of the Yale Gordon Distinguished Teaching Professorship. It’s a little ironic that Carter, who guides companies on ways to navigate new technologies, found himself working with a disruptive technology much closer to home. In early semesters, Carter recommended students use generative AI for extra-credit assignments; he’s now moving to incorporate more of it in the curriculum.

Lyles uses Wolfram|Alpha, a large language model based on mathematics and logics, as a way of having students take over the generative AI reins. Home assignments include using Wolfram for part of the work and understanding what the model can (and cannot) do.

Educators are also using generative AI as assistants for their own work, a result that Stansbury had first visualized. “What’s also interesting to me as a faculty member is that it makes it easier for me to create my own course content, to better design exercises and assignments,” Carter says.

AI FOR STUDENTS

The worries that students will use generative AI tools like ChatGPT to cheat are valid but not justified, according to Stansbury. “Most students that come to college, they want to learn, they want to be educated, they don’t want to cheat, they’re paying to go to school,” Stansbury says.

Lyles cautions against running with a one-size-fits-all approach to AI policy for students. The focus on the humanistic should not be lost in an attempt to embrace tech, Lyles says. “From a teaching perspective, we have learners who are often non-traditional, some are coming back to school as master’s students, or undergraduates who didn’t come directly out of high school,” Lyles says.

Jay Knight, clinical law professor at UBalt, also wants students to take AI with a grain of salt. He reminds them about AI’s shortcomings: The answers are not always accurate, students have to verify them and not accept them as blind facts. Second, using generative AI is not a private endeavor; models chew up the data you feed them. Finally, answers are not consistent—what you receive today might not be the same as one tomorrow.

“ We’re setting up our students for success in society, so we have to recognize that AI is not going away and it’s going to be an AI world.”
DR. JESSICA STANSBURY

Knight teaches students to use generative AI to improve the tone and tenor of their communications with clients. It’s especially useful, he says, for instances where writing is not a strength for students, especially those from underrepresented backgrounds.

The question of equity in access to generative AI tools has also surfaced and is an important one to address. “AI is just a tool, and humans get to decide how it’s deployed, so let’s deploy it for good and help level the playing field for a lot of students especially from minority and underrepresented backgrounds,” Stansbury says.

FUTURE DIRECTION

The data-informed approach to AI adoption is already moving the needle at UBalt, especially by solidifying the University’s thought leadership in advancing AI in education and sparking dialogue within and beyond the campus community.

The University of Baltimore just hosted its first “AI Summit” (p. 18), bringing faculty and students together with local community and industry leaders over three days in June to explore how AI can be integrated into various academic disciplines to drive innovation and excellence.

In September, UBalt joined forces with the University of Maryland, Baltimore County and Johns Hopkins University to launch “AI in Practice,” a monthly webinar series dedicated to discussing the evolving role of Artificial Intelligence in teaching and the essential skills students need to thrive in an AI-enhanced workforce.

And on campus, Stansbury is already looking ahead to working with the Division of Academic Affairs as they collaboratively develop a concrete set of guidelines for faculty related to the adoption of AI, just one of many planned initiatives based on the results of her team’s work with the Ithaka project.

In the meantime, CELTT continues to steer the AI conversation, navigating choppy waters with both enthusiasm and restraint, a tough balance to strike, educators say.

“We have to keep learning about AI, it’s going to continue to grow, and we can’t just stick our heads in the sand and act like it’s not going to happen,” Stansbury says, “with our approach we’re giving students autonomy over their learning, and that’s the way to create lifelong learners.”