A.I. is already deeply entrenched in American schools, and more is on the way. At the most recent iteration of Bank of Montreal’s annual Capital Markets Back to School conference for technology investors, reported panel moderator Jill Barshay, editor at the Hechinger Report, products in the works included anti-plagiarism software, a gamified assessment in which young children read to a computer screen, and a tool for verifying that students in online courses are, in fact, who they say they are.
Still, compared with efforts to envision the future, discussion of what AI offers education right now was decidedly less breathless and more pragmatic.
“Machine learning is great, but I want to think more about human learning,” said TC’s Sandra Okita, Associate Professor of Technology & Education, who studies the use of robots and avatars as “peer learners” that children can sometimes teach as a means of strengthening their own understanding of concepts. Okita added that “a system doesn’t have to be intelligent to help you learn.” But if AI technologies are going to be deployed, she said, it’s important to remember that “tools are just objects, unless used purposefully – the key is what relationship you develop with them. And that’s where learning methodologies such as role play, scripting and dialogical instruction can work really well.”
In a similar vein, Margaret Price, Principal Design Strategist for Microsoft, said that the huge multinational technology company has reoriented itself around the principle of inclusive design. Partly that means doing design both for individuals and “all 7 billion people on the planet,” and partially it means facilitating group interactions so that students can become “digital citizens, collaborators and competitors” who can “thrive in a digital economy in safe, ethical ways.
“We’re looking at co-creation to form hypotheses about creating solutions together,” Price said. “It’s not about novelty or technology for technology’s sake, but about trying to drive student outcomes.”
Stefania Druga, founder of HakIDemia and Afrimakers and creator of Cognimates, a platform for AI education for families, said that with 47 million households having some sort of AI home assistant, some form of literacy is needed for children and their families. Still, she said, in her research, she’s found that children treat AI with less reverence the more they use and master it.
“At first their reaction is, ‘It’s smarter than me,’ but when they learn to use it, it’s not smart.” Druga said she wants to change the terminology from “artificial intelligence” to “extended intelligence.”
Ultimately, Druga said, “the potential of AI in classrooms comes from the diversity of projects that kids can create” – especially projects that allow for age-appropriate expression. One young student she worked with created a program that gave “back-handed compliments” such as “your shoes look so good for being so cheap.”
“For me,” Druga said, “it’s very telling that a ten-year-old wants to teach a robot to be bossy and cheeky.”
— Joe Levine
Speakers quotations may have been edited for clarity.
More from the AI Conference:
- “We Can't Fix Education with Machines:” AI can do wonderful things, but it can’t replace teachers and must promote equity
- “The Future Will Be Nothing Short of Amazing:” For Hod Lipson, the promise of AI significantly outweighs the perils
- A Topic that Pushes Buttons: Sparks fly at TC’s conference on the future of artificial intelligence in education
- AI Can Disrupt Racial Inequity in Schools, or Make it Much Worse: When it comes to tech and education reform, there’s been more talk than transformation