You have /5 articles left.
Sign up for a free account or log in.
Universities are pushing toward more understanding of open-access resources, with the University of New Mexico throwing generative AI into its mix.
Bet_Noire/iStock/Getty Images Plus
Xaver Neumeyer often uses Harvard Business School case studies in his courses centered on finance and innovation at the University of New Mexico’s business school, but he found many of the case studies had the same problem: overlooking smaller, often minority-owned entities.
In the fall, Neumeyer’s students will be tasked with either creating their own case studies or building on existing ones—and will be required to use generative artificial intelligence to do so.
“Whether it was Harvard or other case studies, there are all sorts of entities we typically don’t see covered, so it’s having students examine these things and the entities that are often overlooked,” Neumeyer said. “And it adds another interesting dimension to use GPT, to help put it together or critically evaluate what they’ve written.”
Neumeyer is one of seven faculty members at New Mexico who have spent the summer working to apply generative AI to open educational resources, most often referred to as OER. OER are teaching and learning materials that are openly licensed, adaptable and freely available online.
As the faculty’s eight-week pilot nears an end, each will collect $1,000 stipends as part of the university’s investment into OER, according to Jennifer Jordan, OER librarian at New Mexico. The university also recently received a $2.1 million grant from the U.S. Department of Education to establish an OER consortium in the state.
When Leo Lo, dean of the university’s College of University Libraries and Learning Sciences, first received the grant, he thought pairing OER with AI would help boost faculty awareness about both fields.
“We want people to learn about OER and figured combining AI and OER would be a good way to get people interested and using it as a starting point,” he said. According to a recent OER-focused report from Bay View Analytics, nearly three-quarters (72 percent) of faculty members are “aware” or “very aware” of the resources, with 15 percent of faculty members stating they were unfamiliar with OER. Fewer than a third, though, require OER use in their classes.
At the end of the session, the UNM faculty will compile a guidebook on how to create and use OER, with a chapter dedicated to using AI in OER materials.
“What I want to do with all these programs is to start using and experimenting with it so we know about the current tools,” Lo said. “If they find it’s not super reliable, that’s fine, but we won’t know how to use it unless we actually do.”
The intersection between generative AI and OER exploded with the arrival of ChatGPT in November 2022, according to Christopher Capozzola, senior associate dean for open learning at the Massachusetts Institute of Technology.
“From the first day the commercial tool versions of generative AI hit the scene, this has been sort of beckoning as an opportunity,” he said. “There’s some point down the road when talking about AI and OER would be like telling a fish there’s water in the ocean—but we’re definitely not at that stage yet.”
The legality of what can be used from ChatGPT in OER materials remains murky, according to several experts in the field. Capozzola called it an “untested legal landscape,” while Jeff Seaman, director of Bayview Analytics, called it “potentially problematic.”
Materials in OER, while open, fall under Creative Commons licensing, allowing the material to be reused or remixed as long as proper attribution is given to the creator. Material found on ChatGPT, or other generative AI tools, on the other hand, often does not fall under Creative Commons licensing, and if used for OER, they could violate copyright.
“The question of using AI as a tool with open-source materials is fine,” Seaman said. “Not only fine, but commendable to do things a mere mortal couldn’t do by pulling together material in a new and innovative way, provided it’s only looking at open licensed material. But that last bit is the hard part.”
Dozens of OER experts, including Seaman and Capozzola, emphasized the importance of transparency when creating the open course materials, specifically when citing the use of ChatGPT.
Megan Lowe, director of university libraries at Northwestern State University in Louisiana, also stated the importance of having a human review whatever material ChatGPT produces. She cited several OER materials focused on identifying mushrooms, which, without human intervention, have been misleading readers by identifying poisonous mushrooms as safe to eat.
“I don’t have a problem with using AI; I have a problem with asking GPT to generate something in its entirety and then not reviewing it and doing due diligence,” she said. “I don’t think we can emphasize how dangerous the hallucinations can be, and people laugh it off.”
Seaman said in Bayview’s previous studies, faculty have continuously stated AI’s biggest impact will be on faculty prep and course learning, and “in that case, the New Mexico example is hitting both of those,” he said.
And as both generative AI and OER continue to evolve, higher education can cautiously use both in conjunction with one another.
“As long as we continue our agency as educators, as long as we continue our values around academic integrity, peer review and vetting material using AI tools to improve efficiency, and productivity—why work harder when we can work smarter?” Lowe said. “But you have to have an understanding of the limitations of the current generation of AI tools.”
