Leveraging the finest that GenAI provides

Generative AI became the focal point when over 300 educators and administrators convened to explore, learn, and exchange best practices in pedagogy at Harvard’s annual Professional and Lifelong Learning Summit, which took place in early March.

Bharat Anand, the Vice Provost for Advances in Learning and keynote speaker, focused his discussion on a topic that has frequently arisen over the past year: generative AI and its application across the University to enhance education.

“We’re not observing faculty utilize GenAI to substitute their expertise. They’re predominantly employing it to streamline mundane tasks that often consume precious time.”

Bharat Anand

“This isn’t solely about AI becoming more intelligent,” Anand stated. “What transformed with ChatGPT is its universal accessibility. With a simple text box, anyone — not only programmers — can now harness the capabilities of these tools.”

This accessibility is encouraging experimentation and contemplation throughout Harvard’s classrooms. “We don’t need to await institutional guidance or for conditions to be flawless,” Anand remarked. “Every individual can utilize these tools since the gap between human and machine is closing. That’s why we observe so many individuals beginning to test, investigate, and adapt.”

When tools like ChatGPT were initially made public, faculty reactions were mixed.

“In fall 2023, we experienced both enthusiasm and genuine concern within our community,” Anand observed.

To facilitate exploration, Harvard University Information Technology, guided by the University-wide Generative AI Teaching and Learning working group, initiated a secure GenAI sandbox for Harvard. The objective was straightforward: provide faculty, staff, and students with a space to experiment, prototype, and learn collaboratively.

The outcome? An expanding array of grassroots innovations — along with a novel way to disseminate them.

Learning from each other

Witnessing this interest in experimentation among faculty inspired Anand and his team to connect with faculty across Harvard’s Schools later in the academic year to understand how they had utilized GenAI in their classrooms. One of the most prominent initiatives to emerge from this experimentation is the GenAI Faculty Voices Video Library. This project compiles brief interviews with faculty from various disciplines who have incorporated GenAI into their teaching.

“We aimed for something practical and beneficial,” stated Melissa Tarr, assistant director of programs at VPAL’s Harvard Institute for Learning and Teaching. “These videos offer tangible examples — quick insights that educators can immediately learn from.”

The team requested each participating faculty member to identify a challenge they encountered, a manner in which GenAI was employed to address it, and the insights gained from the experience. Course planning, revising assessments, conducting class discussions, grant editing, and student projects all emerged as areas of innovative activity.

“The faculty we interviewed were open-minded and reflective,” remarked Mary Godfrey, director of multimedia at VPAL. “They utilized GenAI to expand their assignments and test its limits. Several also intentionally demonstrated to students where the tools faltered, as a method to enhance critical thinking.”

It’s about enhancement, not replacement

Use cases encompass generating practice questions, developing tutor bots, providing timely feedback to students, or assisting students in learning prompt engineering.

Anand emphasized an essential takeaway: “We’re not seeing faculty leverage GenAI to replace their expertise. They’re primarily using it to make routine tasks that typically consume valuable time more efficient. Whether it’s synthesizing student responses, responding to inquiries after hours, or creating new practice problems, the goal is to free up energy for more impactful teaching. This mirrors trends in other sectors. The genuine value resides not in surpassing humans, but in ‘automating the ordinary’: in saving time on tasks we already undertake — or neglect — due to time constraints.”

A culture of learning

This ethos — of curiosity, reflection, and shared exploration — has become a hallmark of Harvard’s methodology. Working groups have been active throughout the University, investigating the role of GenAI in research, administration, and pedagogy. The Teaching and Learning working group, chaired by Anand, continues to cultivate resources and monitor developments both within Harvard and externally.

“The video library represents just one component, and we will keep enhancing it,” Anand stated. “We’re striving to support thoughtful experimentation. Concurrently, there is a pressing need for GenAI literacy across the University, ongoing discussions surrounding its ethical implications, and engagement in a more strategic dialogue about what all this signifies for the role of educators.”

Instances from the Faculty Voices Library include:

• Creating new practice problems: Molly Brady (Harvard Law School)
• Enhancing problem-solving abilities: Mitch Weiss (Harvard Business School)
• Course development and syllabus creation: Tari Tan (Harvard Medical School)
• Revamping assessments to gauge learning: Jacob Cook (HBS)
• Introducing concepts at students’ speed: Greg Kestin (Faculty of Arts and Sciences)
• Tutoring bots to assist student learning: Sharad Goel (Harvard Kennedy School)

“The Faculty Voices Video Library showcases the depth and diversity of creativity emerging across Harvard’s campus,” commented Provost John Manning. “I’m eager to witness what further innovation awaits.”

Looking forward: An inclusive conversation across campus

On May 13, Harvard will host its inaugural University-wide symposium discussing the future of generative AI and its consequences for various University operations: research, instruction and learning, administration, and management. Taking place in Klarman Hall at Harvard Business School and supported by a coalition of University offices— including VPAL, the Office of the Vice Provost for Research, the School of Engineering and Applied Sciences, HUIT, the Harvard Library, and FAS — the half-day event will bring forth ideas and inquiries from various fields and schools.

“We’ve noticed significant momentum,” remarked Anand. “Now is an opportune moment to reflect — to learn from one another, to pose challenging questions, and to outline a path ahead.”