Teachers Turn to AI to Make Math Relevant to Students' Lives
Al Rabanera, a math teacher at La Vista High School in Fullerton, California, asked an AI tool to help him teach rate of change to students most likely to tune out abstract formulas. The tool generated an assignment using U.S. Department of Labor data showing how education level and gender affect median weekly income-a topic his students cared about.
When one female student saw the numbers, she had an immediate realization: "Whoa, Mr. Rab! I'm gonna get paid less 'cause I'm a girl?"
That moment captures what educators see as AI's most promising classroom application: making math feel connected to students' actual lives. For teachers like Rabanera, who previously spent hours manually designing such lessons, AI cuts the work significantly.
The Time Problem
Rabanera said creating these types of lessons without AI took him hours of "looking at all the trends and themes, grouping and coding them, and then sharing it back with the kids." AI lets him do it faster.
David Miyashiro, superintendent of the Cajon Valley Union school district in southern California, envisions word problems featuring students' friends' names or homework assignments tailored to individual interests-a baseball fan answering questions about Fernando Tatis Jr.'s home run distance, for example.
"They're learning the same math content, but it's with vocabulary and names of people that they're interested in," Miyashiro said.
Why Student Engagement Matters
More than half of teachers-55%-cite poor student engagement as a significant challenge, according to an EdWeek Research Center survey of 729 educators. Over a third report that students are less engaged in math than in other subjects.
Research shows that linking math concepts to students' interests can make an abstract subject feel relevant. Candace Walkington, a professor at Southern Methodist University, said: "A lot of kids will see math as not relevant, not connected to things that they do, and, as a result, not interesting."
The Reality: Harder Than It Looks
Creating personalized assignments that actually work is more complicated than typing a prompt.
Khan Academy, an early adopter of generative AI in tutoring, removed a feature from its Khanmigo chatbot that incorporated students' personal interests. The tool took too long to respond-more than 5 seconds-and students dropped off. Even when the personalization worked, it didn't improve academic progress or engagement.
Kristen DiCerbo, Khan Academy's chief learning officer, said students see through forced connections. "Kids see through that, and they're like, 'Yeah, I think you're trying to get me to eat my broccoli.'"
When AI Gets the Details Wrong
AI struggles with realism. Walkington, who received a National Science Foundation grant to study personalization in math, found that AI-generated problems often contain impossible scenarios.
An example: a question about a concert where sound reached 400 decibels, a physical impossibility. Another asked students to calculate how many "pins" people wore at different times during a concert-something "nobody would keep track of for any reason," Walkington said.
Some problems are just cringeworthy. Leslie Brown, a 7th-grade math teacher in Texarkana, Arkansas, tested a tool that generated a confusing problem about calculating a donut's circumference while walking laps. Her students also caught an error in a superhero problem: "These boys were like, 'look, Thor's hammer doesn't do that. That's not how it works, Ms. Brown.'"
What's Working Now
Teachers using AI platforms designed for K-12 education-like Brisk and Magic School AI-are finding success by adding choice and personality to assignments.
Rebecca Sheeley, a math teacher in Stockton, California, used Magic School AI to create a Halloween-themed "magic math quest" where students practiced adding and subtracting polynomials. They could choose between a friendly pumpkin patch, a haunted mystery, or a witch's potion lab.
"I could have just given it to them plain," Sheeley said, "but then they just would have been like, 'I'm done,' and not really paid attention to it."
The seasonal twist worked. Students engaged with the assignment because they had agency in how they approached the math.
Next Steps for Schools
Researchers are developing tools that let teachers and students have more control. Burcu Arslan, a research scientist at ETS, is building a tool that lets students pick specific interests-not just "music" but "Freya Skye," not just "sports" but "Boston Bruins." This specificity matters: a music problem can still miss what a student actually cares about.
Teachers remain cautious. They want AI to generate problems that make sense mathematically and in the real world. When AI gets either wrong, the assignment fails.
For now, Rabanera and teachers like him are using AI as a starting point, not a replacement for their judgment. The tool saves time. The teacher ensures the work is sound.
AI Learning Path for Teachers covers practical implementation strategies for educators looking to integrate these tools into their classrooms.
Your membership also unlocks:
