Using AI for student reading skills...and to get my prep period back

Ned Courtemanche | History Department Chair at McDonogh School

Guest Writer

Jul 14, 2025

Get a quick article summary

Image of teacher standing in front of students with computers. AI for students allows them to supercharge their critical thinking and AI for teachers helps save time with lesson plan generation and personalized learning.

At the start of last school year, I came to an uncomfortable realization: that the same AI technology upending the humanities can actually elevate the critical thinking skills that drive the humanities. And while many of my broader concerns about artificial intelligence in our society remain, I can no longer deny that the hours I used to sink into skill work are better-spent utilizing specialized AI tools like Flint. Much has changed in the AI landscape since then, including Flint's sleek new interface and the emergence of increasingly sophisticated multimodal reasoning models. Yet my classroom findings remain consistent: AI tools are force multipliers for teaching critical thinking that can save valuable teacher time while enhancing student ability and experience.

The reading comprehension conundrum

But don’t take my word for it. Let me show you what I’ve learned from two years of tinkering with AI in my classes. The focus of this installment is on reading, that OG skill we’ve struggled to teach since the New England Primer. We all agree that reading is critical, both as a foundational skill and a linchpin for success in today's workplace. Yet many of us in secondary and higher education seem content to leave the heavy lifting to our heroic elementary school colleagues. It’s as if "Fox in Socks" naturally leads students to parsing Madison's Federalist Papers or the painstaking research for that graduate degree in Ecology. Is it any wonder, then, that our media-saturated students find reading both boring and irrelevant or that proficiency scores continue to slump worldwide?

An image showing "Fox in Socks" leading students to Madison's Federalist Papers

I get it - twenty years into teaching, I still catch myself conflating pop quizzes with meaningful skill work. There’s a daily temptation in both the humanities and sciences to jump right into the content of a reading and hope that critical skills will develop with exposure. Even the most enlightened of intentions to teach the sub-skills and strategies that go into effective adult reading run into certain asymptotic limits. The more precise and specialized the critical thinking skill, the more time and energy it takes to effectively teach that skill to classes growing larger by the year. That pop reading quiz might not help me or the student much, but who has the time to run targeted practice work to help my four or five sections of students evaluate something like tone and intention in fictional prose?

AI as reading coach

It doesn’t have to be this way. Specialized AI tools like Flint allow us to build, administer, and assess reading skills like never before. Whether I borrow pre-made activities from the Flint Public Library or simply describe and refine what I need with an AI chatbot in the “Activity Builder”, even five minutes of prep can generate high quality exercises like this Read and Summarize Practice Activity. Just pick a sub-skill or reading strategy - utilizing context clues, pre-reading, visualization, etc. - and you’re a few clicks away. Taking a break from screens in the classroom? Use the “Talk to Flint” chat and brainstorm endless screen-free practice exercises instead. AI is as effective at building formative activities as helping you dream up new approaches to age-old classroom challenges.

A preview image of Ned's Summarization practice - reading comprehension + analysis for middle and high school as AI for students

We talk about “never working harder than your students,” but it’s always been more principle than practice. Let’s say you wanted to help your students with source evaluation to better-differentiate between fact and opinion-based statements while reading. We know skill growth requires deliberate practice which puts you on the hook to build or find a whole bunch of worthy practice problems. How much time would you honestly need to prep this? Thirty minutes? Forty? All for what might be a ten-minute exercise that runs once a year. So much for never working harder than the kids!

It’s an unsustainable use of our prep time and an often-overlooked factor in teacher burnout. Why not flip the ratio and let AI generate those problems to quickly create something like this “Is That A Fact” Activity. Kevin Gregorio has a terrific analogy of using AI as a “pitching machine” for retrieval practice in the classroom. We may always work harder than the students - welcome to teaching - but incorporating AI into my prep allows me to use that time elsewhere without shortcutting the “reps” required for student skill growth.

Quality time with students

Of course, utilizing AI for classroom skill development hasn’t just saved me time, it has drastically enhanced the quality of my time working with students. Just last semester, my Ancient World History freshmen were struggling with a reading comprehension strategy often called “sentence parsing” in which they identify key parts of speech in a paragraph to enhance reading comprehension and efficiency. With Flint guiding them through this Parts of Speech Active Reading Practice Activity, I was free to work the room as a “guide on the side” and humanize the learning - answering questions, refocusing distracted students, cheering on others. I also asked Flint to adjust the difficulty of the problems based on student performance during the activity and could work the room confident that the invisible hand of AI was supporting students with real-time adaptive instruction.

A preview thumbnail of active reading practice for parts of speech, an example of AI for english students

As students wrapped up, I returned to my computer to check their progress using the Flint “Activity Analytics” page. The AI-generated feedback provided a snapshot of the overall class “strengths” and “areas for improvement”, but also the ability to dive into individual student sessions, add my own assessment to compliment the AI-generated feedback, or click on the “Create a Follow Up Activity” to customize future practice activities. Any one of these features extends my classroom reach but, woven together, they unlock transformational learning opportunities. What’s more, it gets me off the computer and back to my favorite part of teaching: working directly with students.

Driving student interest with AI

While this Shangri-la of quick quality prep may have you excited, the prospect of endless practice problems nearly drove my sophomores to strike. Here are the steps I’ve taken since to ensure large language models like Flint also enhance the student experience of skill development. With “reading for fun” seemingly on the decline, I’ve found that it’s important to liven up our skill work by incorporating elements of student choice and gamification where appropriate. Most of my Modern World History sophomores stand to benefit from using active reading strategies. But, being sophomores, they’re also remarkably quick to gripe about using effective protocols like SQ3R. So I will often ask Flint to incorporate student choice into their reading practice to liven up exercises like this SQ3R Practice Activity. Practice turns to empowerment when my sophomores work through AI-identified or generated readings about topics they’re actually interested in.

A screenshot of Ned's SQ3R reading practice using AI for student reading

I also strongly encourage you to play around with the use of gamification in your formative skill work. You don’t have to be a high level programmer or a published mystery writer to build engaging learning games with AI. Just recently, I presented the “Activity Builder” with a challenge - how to improve my student understanding of the early modern English they often encountered in primary sources from Colonial America. With clarifying questions and a bit of testing, this Stamp Act Congress Role Playing Challenge was born which effectively masked critical skill practice in an immersive student experience. There’s a certain alchemic thrill to transforming the grind of deliberate practice into zany classroom favorites like this Context Clues Game Show in the style of “Who Wants to Be a Millionaire”. Giving students the AI tools to build learning games around targeted critical skills is another fan favorite in my classes and usually leads to a Cambrian explosion of students sharing and playing each other’s creations.

There's no time like the present for AI in education

What started as an existential crisis has become one of the most fascinating chapters of my teaching career. AI tools need not diminish the critical thinking we so value in education - they make it more accessible, more engaging, and, frankly, more sustainable for us over-worked teachers. The transformation doesn’t require a complete overhaul in the way we teach, just a willingness to experiment with a few extra clicks in planning that next lesson. Drop me a line with questions, share your journey with the Flint Community, and join the movement - we need more educators exploring the positive educational applications of this disruptive technology.

—

Ned Courtemanche has spent two decades pushing education into the 21st century as a teacher, coach, administrator, and educational consultant. In his current role as History Department Chair at the McDonogh School, Ned has sought sustainable change and updated curricula to build the analytical skills required to help students understand themselves and their digitally-driven world.

As an early adopter of Flint's AI platform, Ned approaches educational technology with equal parts caution, hope, and curiosity. Through this blog series, he documents his ongoing journey of AI-powered learning and teaching, sharing both breakthroughs and challenges for fellow educators navigating this rapidly evolving landscape.