Imagine having personal tutors alongside each student in your classroom. These tutors don’t get tired or grumpy; they speak multiple languages and are patient. If you’re like me, this seems too good to be true. And though artificial intelligence tutors aren’t fully there yet, they are still helpful in my classroom. My journey to the point where AI tutors were essential was making students the center of learning chemistry.

Background and Context

As one of the pioneers of the Flipped Learning Movement and an author of ten books on Flipped Learning, I have seen a lot of change in my career. I started teaching in 1986, and I taught traditionally for the first nineteen years. I stood in front of the room and did the teacher talk thing. Students then took tests and got a grade. In year nineteen, Aaron Sams and I developed the flipped classroom model and saw significant improvements in our student’s test scores. But we weren’t satisfied with flipping our classes, so we experimented with Mastery Learning. We realized this could be the future wave when we implemented Mastery Learning in the early 2000s. However, after our first book, Flip Your Classroom, was published in 2012, we had invitations from all over the world to help schools flip their classes. We then traveled as keynote speakers, workshop leaders, and content creators for many years.

But in 2019, I felt the pull back to the classroom. I had unfinished business. I knew that Mastery/Competency-Based Learning was what I was passionate about, so in the 2019-2020 school year, I went back into the classroom to implement Mastery Learning. Then COVID hit, and things were in turmoil for all of us. I had the chance to help thousands of teachers through COVID with some Flipped Learning principles. But instead of helping as a consultant/speaker, I did it as a fellow teacher struggling alongside the rest of the world’s teachers.

But amid the pandemic, I continued to focus on Mastery Learning. I knew this mode of learning was where we needed to go. So, I continued teaching via Mastery at my suburban high school in Houston. My students were learning at a higher level, and the mastery approach mitigated many of the learning gaps caused by the pandemic. This led to what I believe to be my best work as an author — The Mastery Learning Handbook. The book goes through the details of implementing Mastery Learning in any classroom.

Mastery Learning: Definition and Challenges

In a nutshell, Mastery Learning is the idea that when students “move” through a curriculum, they progress at a flexible pace. When they finish a unit of study, they take an exit assessment, and if they pass, they move on. If they don’t, they get remediation and stay in that unit of study until they learn the content. The idea sounds great, but most teachers hear this and think that the logistics of making that work with 30 kids in six classes a day is impossible. Those logistics were the focus of my most recent book. Besides the logistics, there are other challenges to Mastery Learning. One of those challenges is that teachers don’t have enough time to get to every student. Ideally, more teaching assistants would make the learning even more profound. I have been creative by inviting AP Chemistry students to tutor my first year chemistry students. This works wonders, but most of our school’s AP Chemistry students lead busy lives, and they have limited time to help my students. So what if instead of having AP Chemistry students help my students, I used AI tutors…

The AI Revolution

In November 2022, ChatGPT was released, and the world took notice. Artificial Intelligence has been around for many years, but the release of ChatGPT seemed different. Now, everybody is discussing AI and thinking of ways to utilize it. Most schools fear AI as it can be used to cheat. And though that is a concern of mine, what if we used the power of AI to bring tutors into our rooms to help students with our coursework, especially the difficult concepts?

This past summer, the principal at my school asked me to take the lead in helping our staff think through and implement AI in our classrooms. I spent the summer reading, playing with some AI models, and researching how to use AI in the classroom. I went to a conference, saw a few schools using it well, and realized that AI could be the accelerator for Mastery Learning.

As the summer started, I wasn’t too excited about any one AI tool. I found some great tools that can help my productivity as a teacher, but I didn’t see them as a game changer in my classrooms. But then I got an email from my principal who asked me to check out FlintK12.com. I got on a Zoom call with their founder, Sohan Choudhury, and realized this tool could be the game changer for mastery learning. FlintK12 creates AI activities for students, and when students interact with the AI tutor, they get personalized feedback on their strengths and weaknesses. As the teacher, I can see each student’s strengths and weaknesses, look more holistically, and see what most students struggle with in each class. In the future, the school can aggregate the data on student progress toward identifiable standards.

Below are a few screenshots of some of FlintK12’s activity data. The first image shows an entire class’s strengths after debating the carbon tax, and the second image shows the details of a specific student. Note: These are not my students; they are an exercise done by Flintk12 to showcase their product.

FlintK12 could be a better tool. I love what it does, but some user interfaces could be more intuitive and organized. For example, at this time, I have to copy an “activity” If I want to deploy it to multiple classes. I am now getting to the point where I have created enough “activity” that I would like to have folders to put them in so that my screen isn’t so busy. Even deleting an “activity” wasn’t intuitive. Don’t get me wrong, I like the tool—it is headed in the right direction, and what it does in acting as a tutor is precisely what so many teachers want in AI. I have recently been demonstrating Flintk12 to some teachers that I mentor and, to a person, they want to use it in their classrooms.

I know that FlintK12 is not the only AI tutor, and I am sure that other companies will all be competing in this space, but what excites me is that now we have tools that can take the role of adding additional “helpers” in my room. Remember that, at least right now, I consider the AI tutor equivalent to having a college student in my class helping out. The college student doesn’t know the content as well as I do. The AI tutor sometimes goes off on tangents that aren’t what I want it to focus my student’s attention on. But it is better than not having that “college student” in to help out.

I don’t see AI replacing teachers in a Mastery Classroom. Instead, I see AI tutors augmenting teachers’ instruction and raising the bar on learning. I am incredibly excited about how AI can personalize learning for each student. If you have students whose first language is not English, Flint K12 can talk to the students in their native language. In some tutors, AI can ask students questions about their interests, so the problems skew to their interests. For example, a student interested in airplanes would get physics problems about airplanes, while another student in the same class would get physics questions about cars because their thing is cars.

We are indeed living in exciting times. I am simultaneously fearful and excited about AI in the classroom. However, when I use my mastery learning lens, I only see AI as a way to accelerate the adoption of mastery and competency-based learning.