Students in Miami, Florida, respond to questions about nutrition from NUTRO, described as the “Worlds First Nutrition-Teaching Robot,” in May 1980.
Bettmann/Bettmann Archive / Getty Images
Did you know that AI robots in the classroom transform traditional learning into dynamic, personalized and interactive experiences? That they act as tireless co-teachers that provide real-time feedback, adapt to each student’s unique learning pace and offer a non-intimidating environment for diverse learners and shy students?
We didn’t know that either, until we asked Google’s artificial intelligence chatbot about the benefits of robots in the classroom and got that sales pitch.
Article continues below this ad
We asked the question in response to a now-shelved plan by the Salamanca City School District, in Cattaraugus County south of Buffalo, to introduce “Sally,” a $57,590 humanoid robot, into classrooms this fall. The district planned to deploy it as a teacher’s aide for a small group of students in its STEAM classes – that’s science, technology, engineering, arts and mathematics. The robot would come installed with a district-approved curriculum and strategies intended to foster critical thinking and instructional guidance.
Oh, and there’s a side note: The robot vendor’s parent company also owns a business that makes realistic, AI-powered sex dolls.
After hearing concerns from the community, teachers (and their union) as well as the state Education Department, the district put the plan on hold.
Artificial intelligence seems like a juggernaut barreling its way into our world even as we try to grasp its wide-ranging implications for energy, employment (and unemployment), national security, economics and more — including education.
Article continues below this ad
To be fair, we’ll note that Google’s AI also provided us, when prompted, with a pretty accurate and rather lengthy list of the concerns about robots in school: the safeguarding of student data; the threat to human teachers and to the real, caring relationships they can provide; the blurring and possible expansion of the robot’s role to mimic that of professional educators; the unknown risks to children’s development and mental health.
We’d add another concern that anyone who has used AI can readily appreciate: its unreliability. Artificial intelligence can do a lot, but it also gets things wrong, gives contradictory answers to the same question, and even make things up — known as “hallucinating” — and not even “know” it.
Teachers can make mistakes, too, of course. But it would be dangerous to assume that AI robots come with some assurance of factual infallibility.
Teachers can also think on their feet and navigate many of the nuances of dealing with children, the questions they ask, the issues they bring from home, and the complexities of interpersonal relationships in a class and school environment. No matter how much they may mimic it, robots don’t come with human capacity for empathy, compassion, judgment and adaptability. How are children — who often form a bond with their teachers — to process a relationship with a human-looking mentor with no feeling?
Article continues below this ad
So it’s quite right that the Salamanca district has put this plan on hold. This is a matter that not only the district but the Education Department, the state Board of Regents and the Legislature need to study much more closely — and it underscores that the AI transformation of education isn’t at some point on the horizon; it’s already here.
That’s why the state must accelerate its development of statewide policies for how AI can be used in schools.
How and when should it be used as a teaching tool, and for which students? How will student data be protected? How and when can students use it to help figure out their math homework or complete an essay assignment? How about teaching students AI literacy, and encouraging them to think about the ethical implications that chatbots, design programs and other tools present for human creativity, emotional/psychological well-being, workforce impact and more?
To repeat: School districts are already answering these questions, on their own, because they have to. But the stakes are too high for the state to show up late for this one.
Article continues below this ad
The Education Department says such guidelines are already under development. That’s great; we hope they’re being crafted in consultation with technology experts, educators, child psychologists, parents and students. There should be an ethicist in the mix, too. Maybe even a science fiction writer. And if the Legislature needs to get involved to set rules, lawmakers should make that a priority.
This is not just about introducing a piece of technology into a classroom. It’s about getting students ready for a rapidly changing world.