Since generative AI tools like ChatGPT came into our lives with a bang, teachers have had to cram to keep up with its impact on education. As educators explore ways to introduce K-12 students to the technology, many are wondering how to use it safely and effectively.
Jill Kowalchuk, ’16 BEd, manager of AI Literacy at Amii, the Alberta Machine Intelligence Institute in Edmonton, says it’s important for all of us to understand the benefits and limitations of generative AI, particularly anyone with a stake in tomorrow’s workforce.
“I think now, more than ever, we’re going to need people who understand or have a good AI literacy foundation, whether they’re in a traditional technical role or not in the workplace,” she says.
Here are four things to know about generative AI tools, based on questions Kowalchuk has encountered in her work.
Don’t let fear hold you back
As a former teacher with a masters in education, Kowalchuk understands how AI technologies can promote or inhibit learning. She has heard many of the fears and reservations about its drawbacks.
But she points out that avoiding the technology can lead students to miss out on important learning opportunities. It’s important for teachers and parents to educate themselves so they can help young people navigate this new territory.
Kowalchuk encourages teachers to look for opportunities to learn more. She works on Amii’s non-profit program AI Literacy for the K-12 Classroom, for example, which teaches high school teachers about AI. Amii is running a pilot project to develop a program for elementary teachers, too, and is working to educate parents with its free, on-demand webinar Parenting in the Era of AI. For the general public, Amii offers training on basic AI concepts too.
Sounding good ≠ accuracy
Generative AI tools like ChatGPT are trained to produce coherent, human-sounding responses. Does the quality of the responses correlate with their accuracy?
Not necessarily, Kowalchuk says. Generative AI can produce responses that sound plausible but contain incorrect or even fabricated information.
She explains that ChatGPT and other generative AI tools were developed through a type of AI called machine learning, meaning they were fed huge amounts of information to draw from to mimic human-like outputs. They’re essentially parroting information without actually comprehending the content.
“The computers learn from scraping data off of the internet,” says Kowalchuk. “What they’re doing is recreating text, images and videos based on the material that they’ve learned from.”
“They’re not actually a source of knowledge in the traditional sense.”
She recommends carefully double-checking all output against an authoritative source.
Set some boundaries
Kowalchuk cautions against overreliance on AI. She worries it could lead to cognitive atrophy or the “rubber stamp phenomenon,” where humans reduce their moral agency to accept or reject what AI produces.
She predicts that setting boundaries around the use of AI will become a key issue for education.
It’s important that adults and children foster their critical thinking skills and be equipped to identify increasingly convincing misinformation, she says. Her suggestion for teachers and parents is to introduce generative AI as a form of cognitive extension, much like we use calculators to expand our capabilities.
“We really need to keep human intelligence and cognition in the loop in this process,” she says.
“We don’t want to see a situation where AI created the student’s work and now AI is marking it on the teacher’s behalf.”
It’s an assistant, not the boss
Kowalchuk recommends using the new technologies as collaborators. But always remember you’re the boss.
For example, teachers can use it to create assessment rubrics, lesson plans or newsletters for parents, as long as they double-check the output. She notes that many people may already be using forms of AI without realizing it — for example, for checking grammar.
“ChatGPT is more of a collaborative partner for brainstorming and organizing thoughts,” says Kowalchuk. “It’s really a starting place for the writing process. It’s not the source of truth.”
“I wouldn’t go to generative AI for things that are high stakes.”
We at New Trail welcome your comments. Robust debate and criticism are encouraged, provided it is respectful. We reserve the right to reject comments, images or links that attack ethnicity, nationality, religion, gender or sexual orientation; that include offensive language, threats, spam; are fraudulent or defamatory; infringe on copyright or trademarks; and that just generally aren’t very nice. Discussion is monitored and violation of these guidelines will result in comments being disabled.