Districts around the country may be tempted to follow New York City public schools’ lead in restricting student access to ChatGPT, the artificial intelligence-powered tool that can mimic human writing with eye-popping efficiency.
But they would be making a huge mistake, some experts say.
“I understand the knee-jerk reaction” on the part of the nation’s largest school district, which this week blocked the app on school devices and networks, said Andreas Oranje, the vice president of Assessment and Learning Technology Research and Development at the Educational Testing Service.
The platform “is a new technology that was not part of the standards that they’re trying to meet,” he said. “But it’s a bad idea because ChatGPT is a fact of life. And we want to prepare students for life.”
Instead of squelching students’ access to the application at school, educators need to figure out a way to “create assignments that still get at the skills that you want to teach, but in a way that works with ChatGPT,” Oranje said.
That’s easier said than done, he acknowledged. “I certainly don’t envy their position, especially because these technologies come up so quickly, and really are very disruptive,” he said. “But I also know teachers to be extremely creative people.”
Mary Beth Hertz, a teacher at the Science Leadership Academy at Beeber in Philadelphia, agreed. “My message is basically like, ‘don’t fear this thing,” she said. “‘Like it’s here. Right? You have to embrace it.’”
Ever since OpenAI, a San Francisco-based research laboratory, released the technology late last year, educators have worried about a new kind of high-tech plagiarism. Other educators have scrutinized flaws in the platform, noting, for instance, that it can’t create content based on current events, and can regurgitate factual inaccuracies.
The tool ‘does not build critical thinking skills’
Such potential problems drove New York to limit students’ access to the technology at school, though the district will allow schools to request access to the site for the purposes of teaching students about AI and technology.
“Due to concerns about negative impacts on student learning, and concerns regarding the safety and accuracy of content, access to ChatGPT is restricted on New York City Public Schools’ networks and devices,” said Jenna Lyle, a district spokeswoman. “While the tool may be able to provide quick and easy answers to questions, it does not build critical thinking and problem-solving skills, which are essential for academic and lifelong success.”
It’s not clear how many other districts will—or already have—taken similar steps.
But those that do—or have—should realize that blocking tech like ChatGPT isn’t going to do much to prevent students from using the site, said Joseph South, the chief learning officer for the International Society for Technology in Education.
“You can ban ChatGPT all you want,” he said. “Students are still going to use it. Students have been getting around barriers to cheating forever and they always will.”
In this case, it would take very little tech savvy for a student to access the site outside of school or on a personal computer, and then copy and paste the machine’s work into the district’s official system for submitting assignments, passing it off as their own, South said.
What’s more, if districts block ChatGPT, they’ll almost certainly have to go through the process again soon when the next astonishing AI-powered cheating tool emerges, Hertz added. “It’s like playing whack-a-mole,” she said.
There’s an equity issue here too, Hertz noted. The students unable to access the platform outside school are more likely to come from disadvantaged families who can’t afford devices or internet connectivity.
It might be easier to spot ChatGPT cheaters than teachers think, she added. For one thing, the app rarely includes citations showing where its information is coming from, giving teachers a chance to talk to students about the importance of sourcing and crediting their work.
AI literacy is the new media literacy
Plus, students need to learn how to work with AI for their future careers, said Catharyn Shelton, an assistant professor of educational technology at Northern Arizona University.
For years, the ed-tech world has talked about media literacy in the digital age, she said. That conversation needs to expand.
“We need AI literacy now,” said Shelton, a former high school Spanish teacher. “We need AI literacy so that we can critically think [and] not become idiots and just trust the AI blindly,” instead of using it to our benefit, she said.
For instance, since AI is programmed by humans, racial, gender and other societal biases can wind up embedded in the tech. ChatGPT might be a good way to introduce that concept to students, she said. One potential lesson: Have students ask the tool to compose a love-letter: does it automatically assume that the person’s partner is of the opposite gender?
Teachers could also use the tool to teach editing skills, asking students to turn in the initial essay ChatGPT spits out and then their own, improved version, Hertz suggested.
“What’s scary about this tool is that we have to totally rethink how we’re teaching or what we’re doing in the classroom,” Hertz said. “That’s the world we live in.”
A version of this article appeared in the January 18, 2023 edition of Education Week as New York City Blocks ChatGPT at Schools. Should Other Districts Follow?