The expanding use of AI in public schools is setting off alarm bells in classrooms and kitchens across the country. Teachers are concerned because students are often using AI to complete writing assignments, and they are reading articles describing how their jobs might be displaced because of AI. Parents are concerned because they lived through years of remote instruction and witnessed the disconnection of learning through screens. And both are concerned about student privacy that could be compromised as AI programs soak up personal data on children.

I share those concerns, but my recent experience with an AI bot makes me think AI could have some positive impacts on learning.

At the beginning of this year, my wife was looking up a recipe on her laptop when an ad appeared for an app that promised weight loss through intermittent fasting. We had both used intermittent fasting to lose weight gained over the holidays. The method we used required severely limiting calories two days a week while eating with abandon on the other days. That protocol helped us for a short time, but we both disliked tracking calories and, consequently, after a few weeks we returned to our normal diets, and our higher weights.

This new app, however, offered a different regimen: we would fast for 16 hours a day and during an 8-hour time interval, consume whatever we wanted. The app also provided a “coach” that promised to assess our daily diet and offer encouragement to stick with the intermittent fasting regimen if we got discouraged. For $5 a month, my wife figured it was worth a try and, since she prepares most of the meals, I became a participant.

The next day, we met Coach Avo, who became the third member of our family. He joined us before we ate to rate the meal my wife photographed for him, presenting her with an easy-to-understand chart and offering her specific recommendations for how to make the meal “optimal,” the highest rating possible. Coach Avo also sent periodic texts and emails to my wife, reminding her to stay hydrated, eat mindfully, get some exercise; tracking her eating timetable; and, encouraging her to stay the course with the fasting even on days when she gained few ounces of weight.

Coach Avo’s invariably uses skillful speech. He never criticizes my wife’s food choices or chastises her if she fails to adhere to the app’s recommended schedule. Instead of ridicule, Avo offers cheerful reminders that staying on schedule would help her achieve her stated weight-loss goal and avoiding sugary foods would enable her to sustain the fast without having cravings.

Most crucially, Coach Avo measures my wife’s progress based on the long-term weight loss goals that she set. He didn’t judge her based on a pre-determined, rigid, one-size-fits-all diet based on a fixed calorie count. Nor did he judge her on her adherence to the eating intervals recommended by the app. Instead, he kept her long-term goal at the forefront, nudged her to avoid foods that make her want to eat more often, and reminded her constantly that the body burns the most fat from our bodies it in the last hours of the fast. In that way, Coach Avo was is gradually changing her eating habits, and mine.

My wife and I know that Coach Avo is an AI bot. We know his ability to interpret photographs of our meals, his “tips” on what to substitute to make our meals more satisfying, and the data he presents in his motivational charts are all generated by algorithms. We also know that Coach Avo has no sentience. He really doesn’t care if we follow his advice, or if we achieve our goals, or if he is connecting with us. His “tips” and “pep talks,” like his “knowledge,” are all algorithmic. Even though we know Avo is inanimate, we both feel better when he gives the meals we prepare an “optimal” rating, when he gives my wife positive feedback on her careful tracking of meals, hydration and exercise, and when her daily weigh-ins show that she is accomplishing her ultimate goal.

Coach Avo’s approach works because he doesn’t evaluate my wife by comparing her rate of weight loss to an age cohort, or to a cohort of individuals seeking to lose the same weight she hopes to lose, or to any set of means-based standards. He evaluates her based on her own goals and advises her to make better choices in her diet based on research, knowing that there is no one best way that works for everyone. Coach Avo’s approach works because he communicates that weight loss, like learning, can only occur when an individual is committed to it.

As one who spent decades working in public education, I believe coaches like Avo could help in schools, especially if schools used AI to abandon the rigid standardization used to measure students and replaced it with the kind of individualization that might be possible with AI. If schools measured student performance based on the learning goals they set for themselves instead of standards based on the rate of learning of students who are the same age, a seismic shift in education would take place. Instead of time being standard and learning being variable, learning would be standard, and time would be variable. In such a paradigm, teachers would need to help students set realistic and attainable learning goals for themselves and guide them during the time it takes for them to reach those goals.

In such a paradigm, the metrics used to reinforce competition and sorting-and-selecting based on standardized tests would fall by the wayside. In such a paradigm, AI would be used to help teachers motivate students to reach their own goals. In such a paradigm, AI could transform schooling instead of making a system developed in 1920 more “efficient.”

Wayne Gersen is a retired public school administrator. He lives in Etna.