Is There Such a Thing as 'Good AI' in the Classroom?

When we think of “artificial intelligence,” we conjure up all kinds of images—from Ex Machina-like robots to chatty and sometimes obtrusive virtual assistants like Siri. Nerdy and flawed, these clever bots are finding their way into our imaginations, our pockets, our homes, our cars, our work, and yes, our classrooms. No, not our classrooms! This image gives many of us pause—after all, classrooms are nurturing, human-centered physical spaces where the bond between a teacher and a student is sacred.

As jarring as the idea may seem initially, it has a lot of potency when you look at AI as a means through which we can amplify the teacher’s impact on the lives of each and every one of their students.

AI as an amplifier rather than a teacher replacement is an idea we can get behind, right? Potentially. But not if the amplifier acts as a stand-in for social-emotional learning, or if the time we have set aside for authentic social discourse among 30 humans is replaced by device-mediated and isolating learning activities. I think we can all agree that AI that separates a teacher from students is bad AI.

So, what could be “good AI” in the classroom?

In Grades K-3, teachers spend a lot of time doing one-on-one observational assessment of oral fluency. It requires a timer and means for capturing words spoken correctly and incorrectly—one of those vital teacher activities during the period of time when children are learning to read. What if a teacher could deploy an intelligent assistant that recognizes the speech of a child—like Alexa does for us adults at home—and it could assess the child’s oral fluency, provide the teacher with a report, and then allow the child to “practice” and improve their reading skills? In short, it would be an intelligent assistant that saves teachers time and gives more opportunities for the child to practice their reading.

At HMH, we consider that not just good AI for a teacher and a student, but game-changing AI. That’s why today, HMH announced a new partnership with Amira Learning, which has created an AI-powered technology solution that does exactly that.

Now, think beyond the point where students are learning how to read. Think about the classroom where the kids are reading. There is an American flag in the corner, a chalkboard (or better yet a “smartboard”), and several computers along one of the four walls of the classroom. There are 30 kids—some working together on projects, some working with their teacher on a learning activity, some reading quietly at their desks. While it stands to reason that all of the children may be around the same age because they are in the same classroom, we know that no two children are alike when it comes to what they know and what they are ready to learn. It’s a class of 30 snowflakes. That’s what makes them unique and very special. It is also what makes a teacher’s job incredibly challenging—first assessing what each child knows, and then ensuring each gets the personal instruction needed to advance their learning.

Again, what if an intelligent assistant could assess those students for the teacher, tell the teacher what each student knows, and then provide a personalized recommendation for engaging instructional activity for each student? Or, if the teacher prefers, what if the intelligent assistant could go ahead and provide an engaging instructional activity directly to the student? And, as the student learns about math, what if the intelligent assistant could learn more and more about how the student learns and provide those insights to the teacher? Earlier this year, we welcomed award-winning adaptive learning solution Waggle to the HMH family to make this scenario a reality.

At HMH, we think that is good, game-changing AI.

Those two examples of “good AI” are what we call “purposeful technology” at HMH—the no-empty-calories version of technology in the classroom that creates capacity for a teacher and frees them up to do high-impact work. It’s the kind of technology that tracks student progress while the student is learning without the teacher having to administer and grade a battery of tests—the kind of technology that can help teachers differentiate and even personalize their instruction.

Increasingly, that purposeful technology can and will be powered by AI. Because while a bot is incapable of determining whether a child is hungry or sad or both, teachers are incapable of reasoning on a petabyte of anonymous student data to find just the right intervention at the right time for each one of their 30 very special students.

Good AI allows the teacher to remain at the center of the classroom. It helps boost a teacher’s superpowers behind the scenes. And that’s game changing.