Florida universities look to stave off AI in academic malfeasance
TALLAHASSEE - As artificial intelligence becomes an increasingly high-profile topic in higher education, Florida universities are holding discussions about how to harness the technology's power - and stave off the potential for academic malfeasance.
The University of Florida Board of Trustees, for example, received an update Thursday from the school's Provost Joe Glover on artificial intelligence, or AI, initiatives at UF.
The state's flagship university is home to a supercomputer known as HiPerGator, which was a $50 million gift to the school from NVIDIA, a Silicon Valley-based tech firm, and company co-founder Chris Malachowsky, a University of Florida graduate. Touted for its artificial intelligence capabilities, the supercomputer is being used by researchers from UF and other state universities to tackle pressing agricultural and environmental problems.
But Glover said "we need to pay attention to" another "very important development" in AI - a technology that's known as generative AI. Generative AI has exploded into the tech world in recent months and drawn widespread attention for its ability to generate things such as text and images.
"This is an important new technology. It will totally disrupt higher education," Glover told the trustees.
The provost pointed to what he called an "emerging research agenda" in generative AI.
"As everyone knows, generative AI hallucinates. ChatGPT makes mistakes, it doesn't give the right answers. It is subject to flights of fancy and depression," Glover said, referencing a widely used generative AI tool.
OpenAI, the artificial intelligence research company that developed ChatGPT, describes its text models as "advanced language processing tools that can generate, classify, and summarize text with high levels of coherence and accuracy."
Human efforts are needed to safeguard against ChatGPT's raw capabilities, Glover said.
"And so, it needs a validation ecosystem, which people are working on now. It needs development and collaboration with subject-matter experts. It needs ethics, security and policy built around it," he added.
Meanwhile, UF continues to pump significant resources into its AI initiatives. During a March 16 trustees meeting, Glover described efforts by the school as "building an AI University and infusing AI into everything that UF is doing."
The school also has proposed a new degree, the Master of Science in Artificial Intelligence Systems, which would be established within the Herbert Wertheim College of Engineering.
Florida Gulf Coast University's Board of Trustees is also eyeing the issue. The board next week will hold a discussion on artificial intelligence and its "threats and opportunities to teaching and learning."
Trustees are scheduled to hear a presentation on technology tools such as ChatGPT during a board meeting Tuesday. A discussion about students' use of generative AI, particularly in assignments that involve writing, is part of the meeting's agenda.
"Following Spring Semester, 2023, FGCU saw no increase in cases of student misconduct overall or cases of Academic Integrity specifically compared to the previous year. However, anecdotal evidence from instructions of English Composition suggests that students are using ChatGPT specifically in writing assignments," a description of the planned discussion said.
The university has a license to use the plagiarism-detection software TurnItIn.com, which the school said includes an application to detect "signatures of AI-generated prose." Professors currently are using it to screen students' work.
"Moving forward, the Dean of Student's office will begin tagging academic misconduct cases for instances of Generative AI usage, and an explicit syllabus statement will be generated for faculty use to prevent cases of unauthorized assistance," Tuesday's agenda item said.
FGCU professors have been using generative AI "with excellent results," according to the meeting item.
"On the positive side, Generative AI, whether used for writing, art, graphics or other applications is likely to be a transformative innovation in higher education," the agenda said.
Glover also predicted that generative AI could reshape the internet as the technology's use becomes more widespread.
"There is going to be so much content generated on these generative AIs, that it will flood the web, and it will become unreliable and untrustworthy. You'll have some generated by people on the right, some generated by people on the left. Some promoting vaccines, some by anti-vaxxers. And so, people will lose confidence in it," he said.
