AI: One Size No Longer Fits All

New, flexible technology is allowing educators to tailor learning to the needs of each student, while giving instructors more free time. But there may be a catch.

I remember trying to learn math in elementary school. I was a natural when it came to words—I can't remember when I didn't know how to read—but numbers were another story.

My teachers had a singular way of drilling me. They'd make me stand at my desk (while all the other children watched) and look at the problem on the board, something mind-numbing like 7x5, and I would stare at it trying to remember my multiplication tables, which just never seemed to stick, until I would angrily hear, "Okay, just sit down."

And I would sit down. It was how things were done.

Another favorite method: After recess to get back into class we'd each have to stand in the hall while the teacher held up flash cards with a problem on them, which we had to solve before we could get back into class. I was often the last one left standing. It was not a situation where being the last one left standing was a good thing.

I eventually learned my multiplication tables, of course, and in fact took AP Math in high school and made it—barely—to Calculus II somewhere, where I utterly collapsed. Eventually the ideas just refused to penetrate my brain. The irony was I was a good student in English and other languages, even though I rarely studied those subjects, and not so great in math, where I worked incredibly hard just to keep my head above water. My teachers seemed to think I just didn't care. They were wrong.

I have to admit the experience left a strong dislike for math in my soul. Many times I've wondered how my relationship with all the wonderful things math opens up—science, music, economics, architecture—would have been different with AI.

I went to school back in the Late Mesozoic, when it was generally believed there was one ideal way to learn (or, if more than one, not very many and they were rather similar). Now, I have no doubt my teachers had my best interests at heart. But often learning was like running a brick wall, picking myself up, and trying again. Today is a completely different story: Children can learn in a variety of ways tailored to them specifically, through AI that monitors their progress and generates instruction based on what they need. Platforms can track student progress from kindergarten all the way through high school, painting a picture of students' progress that is much clearer than the more vague records that follow them through the years and grades and institutions now. The software can determine what educational approaches work and what don't—something I wish existed when I was in colds sweats over 50471÷12.

Furthermore, AI is proving to be a powerful tool for enhancing student engagement. By generating tailored content that resonates with their needs and progress, it can keep students more interactive and committed to the learning process. AI-powered tools can create dynamic and interactive learning experiences, such as gamified lessons, virtual reality simulations, and augmented reality applications that bring abstract concepts like math to life.

Paired with VR and AR technologies, AI can simulate realistic environments for experiential learning, such as recreating historical events or scientific phenomena, adding depth to theoretical knowledge. Randi Williams, PhD, a program manager at the Algorithmic Justice League, observed playful interactions between children and robots in her studies. Dr. Williams says the children not only tried to learn from the agents, which of course is expected, but also teach them. True interaction.

AI can give overworked teachers relief, generate lesson plans and templates, and individually monitor students' unique progress from K all the way to that final walk down teh aisle. The software is smart enough to grade the teachers' native tests...no need to remake exams or class lessons.

There are potential dangers, as there are with everything. It's important to understand that this technology, all-encompassing as it is, is not a substitute for teacher-student interaction. And AI should also not be used—in this writer's opinion at least—to tailor lessons to students to too large a degree. Let's face it, the real world has certain parameters, and it's important students learn to swim in this sea, even if it's more difficult for some than for others. AI may be accommodating, but reality often is not.

To that end, it's important that educators and administrators don't use AI for a high tech baby sitter, something for which it is not intended. It should stimulate creativity and curiosity, not replace it. Children need to learn early on that this is not a magic "answer machine." So while using AI to uncover sound and unsound reasoning in an argument is good usage, simply believing what the text says is not. While exploring out-of-the-box topics is an excellent use of AI, not having students independently read up on their new discoveries from multiple sources is not.

I also have a concern that very young children may not fully understand they are not interacting with a real person. It's all too easy to imagine them seeing a bot or agent as an extension of their teacher, the way a small child may believe their world of cartoon friends to be real.

Some recent studies have found that even adults can become attached to their AI agents. The movie Her was about exactly that. Now picture a highly-imaginative seven or eight year old, whose ideas of the world are still forming, using an AI trainer, mentor or friend.

And there's the cost barrier. Relatively well-funded districts can absorb the cost of an AI-enhanced curriculum better than poorer schools. Of course it's the poorer schools that would arguably most benefit, where educators and their resources are likely stretched thinnest. Time will democratize the technology, the way it has almost everything, but right now some of the benefits of AI can be prohibitively expensive.

And there are some wrong ways to view AI. It is not a way to cut costs by turning tasks over to automation while employing fewer human teachers, or less-qualified teachers. Like everything else, AI is only as good as its inputs. As Tom Ryan, co-founder of the K-12 Strategic Technology Advisory Group, said in a recent panel discussion, "We should be looking at how to increase efficiency with AI so we have more money to pay and train teachers."

Could Orville and Wilbur have imagined the future when the first Wright Flyer left the ground for 12 seconds and flew 120 feet? AI is the most game-changing technology in the last 100 years. Yet its biggest impact will be in ways we have not yet begun to imagine. This is how technologies inevitably develop: They surprise even their inventors. Just ask Alexander Fleming, Henry Ford, or Thomas Edison.