Colleges and universities are rushing to implement AI technology across their IT infrastructure, and higher ed’s enthusiasm may be greatest around the technology’s potential to revolutionize teaching and learning. While some state flagship universities and well-endowed private colleges can build in-house models, smaller schools are working with vendors to offer cutting-edge assistance.
California Lutheran University recently unveiled its enterprise-wide AI solution, Ask Gumby, to provide students personalized guidance on scheduling and assistance with course materials. Entrinsik, which has worked with the private faith-based university for over a decade to build a custom suite of data analytics and business intelligence solutions, integrated its chatbot across its portal and learning management systems at little or no cost.
“Larger R1 institutions like the University of Michigan have been able to build AI solutions like Maizey in-house. Smaller institutions don’t have those in-house resources,” says Zareh Marselian, CIO of California Lutheran.
One of Ask Gumby’s greatest advantages is providing students with one-on-one tutoring. For example, faculty can feed the AI with course materials to reinforce concepts students are struggling with on homework or a quiz. Direct access to student queries gives faculty a quantifiable metric on the concepts they need to reinforce in class.
More from UB: The 20 ‘New Ivies’ that are capturing employers’ attention
“One thing that is very unrealistic is finding a teaching assistant at an odd hour waiting to help you out,” Marselian says.
Beyond curriculum, Ask Gumby can also identify at-risk students by compiling data on class attendance and their visits to the wellness center, says Madhavi Chandra, chief product officer at Entrinsik. “You can take curated information that already lives in the institution and use that to come up with a proactive way of alerting an advisor, a professor and even the student.”
The Center for Empowerment and Pedagogical Innovation at San Francisco Bay University is working with Alma Learning to train faculty on how they can use an AI assistant to refine their course design for neurodivergent and foreign students and provide trauma-informed teaching.
“An assignment rubric can become a living thing,” says Anzar Khali, chief learning officer at San Francisco Bay. “If a rubric demands a student share an example of a successful leadership figure from the 19th century, [the AI] can suggest that a student may be unsure what ‘success’ means,” depending on their cultural upbringing.”
AI both institutions, AI assistants will give students running feedback on their proficiency ahead of highly consequential midterms or final exams.
“If students don’t get quality, good, consistent feedback, they’re very likely to not know where they are,” Khali. “That uncertainty can potentially cause mental pressure, depression and stop them from submitting future assignments.”
Why faculty outrank AI assistants
While AI assistants can help students around the clock, faculty will remain the experts on course content. San Francisco Bay faculty must design a robust course for the AI assistant before it can suggest more accessible curriculum and pedagogy.
“If that is not done properly, you can’t deliver universal learning design,” Khali says.
Marselian at California Lutheran says Ask Gumby may provide wrong answers even though it is trained to assist students based strictly on course material. Faculty must monitor the AI assistant’s responses and correct them.
“I’m not blinded by the shortcomings or the hype surrounding gen AI,” Marselian says. “Every technological advance has its drawbacks and disillusionment. I’m not counting on AI’s future promises, of which there are many. We are maximizing what it can do for our students today.”
