Teacher trainer Mark Hines was conducting a workshop for about 100 Hawaii public school educators in April when he did an informal poll: How many of you are using artificial intelligence in any way?
“Almost no one raised their hands,” recalled Hines, who is founder and director of the Kupu Hou Academy, a teacher professional-development program at Mid-Pacific Institute.
Hines then posed the same question to the high school students at the back of the room who were serving as helpers at the workshop.
“They all raised their hands,” he said.
The moment hinted at some of the massive problems in education that have arisen with the recent explosion in public use of generative AI — intuitive platforms that people of nearly any age or ability can easily prompt to produce content. The dramatic shift has rattled many public and private K-12 schools and colleges in Hawaii and around the globe, especially since the launch of ChatGPT in November.
While many educators and students are embracing AI and looking for ways to leverage it to elevate education, others are arguing it is hurting students by giving them a tool many are using to cheat and short-circuiting their learning to think for themselves.
Educators last school year were already reporting rising AI plagiarism — copying and pasting the content an AI chatbot produces. Generative AI platforms can now almost instantly produce humanlike text compositions, as well as images, math solutions, coding, videos, artwork, music and more, and that is creating ethical dilemmas, growth opportunities and privacy concerns for both students and educators.
Still, no sweeping bans on AI are being considered by the state Department of Education or the University of Hawaii, as the islands’ public schools resume classes starting Monday, and UH and most private schools follow suit this month.
Instead, both state agencies as well as many of Hawaii’s leading private schools are favoring finding ways to teach about and work with AI.
“Generative AI is not going to go away if we sort of bury our heads in the sand and ban its use across the university in a blanket way,” said Gloria Niles, director of online learning for the 10-campus UH system.
>> RELATED: Ask ChatGPT: Is a student who uses AI cheating?
>> RELATED: Teachers, students divided over AI’s risks, benefits
However, while students are tending to dive enthusiastically into generative AI to see what it can do for their schoolwork and lives, surveys suggest educators are lagging far behind, with very few well prepared to teach about, and with, fast-evolving AI technology.
ONE RECENT survey found that 87% of K-12 educators in the U.S. had not received any professional development training on how to incorporate generative AI into their work.
The survey of teachers, principals and district leaders conducted in May and June by the Bethesda, Md.-based EdWeek Research Center also found that only about 10% of educators nationwide feel they know enough about artificial intelligence to teach about it or use it in some way in their work.
Those percentages likely are about the same among educators across Hawaii, key stakeholders say.
“From my understanding, there were very few (Hawaii) schools and organizations that even really acknowledged and/or discussed ChatGPT in the spring last school year … , ” said Brian Grantham, director of educational technology at Mid-Pacific.
“I knew of very few schools who actually did any sort of training or (professional development) or any sort of informational sessions about what this is and how it’s going to be used, or even coming out and making a statement about, you know, the stance of the school (on) can it be used or can it not be used,” said Grantham, who also works alongside the school’s Kupu Hou Academy as it provides professional development to teachers from public and private schools.
“I’m pretty sure that the majority of our kids across the island went through the spring with no guidance at all,” he said.
If a survey like EdWeek’s had been done of UH faculty back in the spring, a similarly small fraction probably would have said they knew enough about generative AI to teach about or with it, said Niles. She thinks at least some faculty will have taken the initiative to learn about AI over the summer, but some are resisting.
“What I’ve seen is that faculty who are adamantly opposed to it … have not tried generative AI themselves,” Niles said.
She is encouraging faculty members to experiment and even “be almost vulnerable with their students — that this is new, we’re all learning this. … And have these discussions with their students about what are their students’ ideas about how they could use this to make their learning more efficient? How can we create this assignment that uses hard human intelligence as well as leverages the artificial intelligence of generative AI?”
AI PLAGIARISM, other types of shortcuts and academic dishonesty by students are among the biggest worries associated with generative AI in education.
One new survey suggests more than 4 in every 10 teenagers are likely to use AI to do their schoolwork instead of doing it themselves this school year.
The survey of 1,006 13- to 17-year-olds nationwide, conducted by the research firm Big Village in July for the nonprofit Junior Achievement, also finds that even 60% of the teens surveyed said they consider using AI for schoolwork as cheating.
An informal poll in January by the Stanford University newspaper found that 17% of 4,497 student respondents there had used ChatGPT on their final exams — even while the school had forbidden its use.
At Lahainaluna High School on Maui, English teacher Jarrett Chapin said he was stunned in the spring by the number of students trying to pass off AI-generated content as their own work. He recognized it mainly through meticulous reading, because often online platforms created to spot AI content “weren’t up to the challenge,” he said.
Private schools were hit as well. ‘Iolani School, for instance, “did see cases in the spring semester of 2023. We adjudicated them in a manner aligned with our longstanding standards for academic honesty and our student handbook,” Head of School Timothy Cottrell said in an email response to questions from the Honolulu Star-Advertiser.
Some educators also fear that generative AI will make it easier for students to skip whole chunks of learning, and that students will assume that flawed AI content containing factual errors or biases is accurate. ChatGPT, which draws from billions of pages of data across the internet to craft its responses, itself carries a disclaimer at the bottom of the website that says it “may produce inaccurate information about people, places, or facts.”
YET OFFICIALS with Hawaii’s public school and public university systems say that flatly prohibiting generative AI from all academic work would be the wrong tack.
At DOE, a work group is being formed to focus on understanding and appropriately using AI. “We want to make sure that everyone from our students and staff are able to use it productively,” said Winston Sakurai, curriculum innovation director for the DOE Office of Curriculum and Instructional Design.
He sees AI as the latest evolution of tech tools that initially met pushback but eventually became accepted for use by students, such as the calculator, spelling and grammar checkers, Google search, and digital assistants such as Alexa and Siri.
Teachers, too, can benefit from AI, he said, using it to help design lesson plans, saving time while making curriculum richer.
DOE last week was set to distribute guidance on AI use for its statewide system of 258 traditional public schools with their 157,000 students. “The key is understanding ethical use of technology,” Sakurai said.
Meanwhile, at UH, “we’re recommending to each of the campuses that every instructor needs to have a statement in their syllabus … of what are the expectations for students in that particular class,” Niles said. “It could be anywhere from if the faculty wants to ban it and it can’t be used at all, it’s considered academic dishonesty, to faculty who are requiring the use of generative AI in certain assignments.”
UH also plans to launch systemwide online professional development sessions across the fall semester “to help faculty to learn and to think about how can they re-frame their (student) assessments,” Niles added. “I often say to faculty that if generative AI can do your assignment, you need to rethink about what’s the learning objective of that assignment.”
PUNAHOU, ‘IOLANI and Kamehameha schools and Mid-Pacific Institute also told the Star-Advertiser that they will not impose schoolwide prohibitions on AI and instead are actively exploring its use.
At ‘Iolani, “we recognize that this is a pivotal point in human history and that AI is here to stay,” Cottrell said. “While we aren’t becoming an ‘AI School,’ we are committed to learning about AI, staying at the forefront of technological development, and thoughtfully implementing its use when and where it is appropriate and valuable.”
Over the summer at ‘Iolani, more than 50 faculty have “engaged in targeted AI learning and professional development projects,” Cottrell said. A custom portal to ChatGPT for students is being built, a new “AI educator” on staff will lead a program providing professional development for public school educators, cyber safety and anti-bullying policies and procedures are coming, and AI is being integrated into media literacy courses, he added.
Punahou also sees the introduction of generative AI as a “watershed moment,” Robert Gelber, director of communications, said via email.
“AI platforms can be powerful tools, and we need to prepare students to engage with them as they will undoubtedly transform the workplace. We are also committed to intellectual and academic integrity. We aim to deepen our teaching with our students about critical thinking, responsible authorship, and the importance of citing sources.”