No one would remember the line, “Keep being good and special, Ponyboy. Don’t change.” It just doesn’t hit the same as S.E. Hinton’s classic line, “Stay gold, Ponyboy, stay gold.” 

Many English teachers, myself included, would consider it an atrocity to diminish this canonical line in literary history to such sterile and lifeless simplicity. Students deserve a better reading experience than that. Yet this is just what happens when teachers decide to level a text down to a student’s “instructional level” rather than scaffolding up their access to a complex text. 

In my own district, just two years ago, educators raved about a new AI tool that helps teachers level text and encouraged the department to spend thousands of dollars on it. When I first heard about it, I admit, I was dazzled. The most time-consuming and work-intensive part of teaching for me has always been differentiating for the small percentage of kids who just don’t get it the first, second, or even third time you teach something. 

This program was going to do it all for me, but the caveat was that it would also do all the thinking for the kids. When we keep kids from actually doing work that requires effortful thinking, they don’t learn anything at all. 

While many states are passing legislation that focuses on the Science of Reading, moving districts away from programs that promote text leveling, many edtech companies are promoting this practice through their newest AI models. In our world of quick and snappy AI solutions, the leveled text shortcut is all too easy to take, but we should not give in to temptation.   

The History of Text Leveling

The practice of teaching students at their “instructional” level has a long, storied history leading back to the founding colonies. If that sort of history interests you, I highly suggest checking out Timothy Shanahan’s “Leveled Reading: Leveled Lives,” which, for me, was one of the most eye-opening takes on reading education in this country since Natalie Wexler’s “The Knowledge Gap.”

Essentially, the practice of leveling text rests on the belief that we should give students texts that are at their “instructional” level only; for example, if a seventh grader is significantly behind, they may be reading on a third-grade level, so as their teacher, I should teach them using only third-grade-level books. This practice was popularized by elementary programs that assigned students reading levels such as the Fountas and Pinnell levels, Accelerated Reader, and Guided Reading. Eventually, it pervaded secondary intervention spaces and general education classrooms. 

The problem with leveling text is that decades of research, more recently popularized by Science of Reading advocates, show that it doesn’t give students much room to grow. 

What students need to grow is complex texts, but not only that, complex texts with a teacher modeling how a good reader approaches these texts, while also scaffolding access to them through chunking, discussion, vocabulary practices and all the critical components that we know make up good teaching. 

As much as research pointed toward the necessity of complex texts, the pressure to maintain instructional levels was a tension I had to navigate daily, and that tension came to a head when my district adopted a new curriculum.

The Faustian Bargain

“It's too hard and too boring. Our dyslexic and ADHD students will never be able to read Shakespeare, Chaucer, or Remarque.” 

I heard comments such as these far too often when my own district in New Jersey switched to complex grade-level texts. As a literacy coach in 2025, I lived in the middle, translating the will of state- and district-level administrators to the classroom and the average teacher. 

Surprisingly, I got the most pushback from eighth-grade teachers who were terrified that even the general education students couldn’t handle reading the classic anti-war novel, “All Quiet on the Western Front.” That is when Diffit was first suggested to me. Diffit’s primary function is to level texts; it even promotes itself on Google as “an AI tool teachers use to get ‘just-right’ instructional materials.” 

In the midst of this difficult year implementing a new curriculum, our teachers were unfamiliar with evidence-based practices and used to a home-brewed curriculum; students were also not used to the rigor of it all, especially the texts. 

One group of teachers who struggled was the special education teachers, who, admittedly, had too many preps and too large a caseload to adequately understand these new mandates. It was then that a well-respected Learning Disabilities Teaching Consultant (LDTC) in the district first suggested I check out Diffit and start putting each chapter of “All Quiet on the Western Front” into this AI program to give eighth-grade special education students a third-grade-level summary. 

Initially, I was impressed by the power of this new AI program, but I also felt conflicted because my supervisor had just told me about Shanahan's work. The world didn’t freeze, per se, but in that moment, it certainly felt like I was living in “The Matrix” — the electronic alternative universe popularized in the movie of the same name. 

“Isn’t that like…frowned upon? That’s not good practice, right? I thought we don’t level texts,” I asked my supervisor cautiously. 

My gut and everything I had learned about teaching reading told me this was wrong, but this was a well-respected educator with decades of experience working with students with disabilities; surely she knew better than I did. I figured that since I was a general education teacher, I was clearly missing something about teaching reading specifically to students with disabilities, so I trusted the veteran educator at the time. 

The Fallout

After taking the veteran teacher’s advice, I recommended text leveling to the teachers I coached, and even used the site myself to give a dyslexic, ADHD student I tutored a shortcut. The results were lackluster. My student didn’t show any interest, would wait for me to tell her the answers, and didn’t remember the content of the text from one session to the next. Her special education teacher wasn’t reporting any improvement in class from this resource, either. 

It didn’t take me long to shift gears back to what I knew would help. Reading and re-reading a text (yes, the real one); mapping out the sequence and character; discussing the theme; exploring vocabulary; and building background knowledge. All of this is done with intentionally spaced and interleaved retrieval practice to ensure maximum commitment to long-term memory. 

Once my student began working to understand the dense text, the payoff was immediate. Our discussions of Paul Bäumer’s psychology and the trauma of war were profound. This dyslexic, eighth-grade student with ADHD, whom many had already given up on, started to join her peers in these discussions back in the general education classroom. My student didn’t need a different text; she just needed a different roadmap and timeframe to get there. 

Three years later, those same teachers who doubted this student could meaningfully engage with such a complex text are no longer surprised when “All Quiet on the Western Front” is their students’ favorite read of the year. Why? Because it deals with real-world topics in a mature way that middle schoolers are often shielded from. 

Student success with the challenging text, despite the availability of powerful, easy-to-access AI, is part of the larger conversation about the quality of the tech tools we use and how they may or may not contribute to more meaningful learning experiences. 

The fault lies not in the tool itself, but in how we train ourselves as educators to misuse it. 

The Redemption

Faustus' tragedy was not that he sought power, but that he ultimately lost control of the force meant to do his bidding. Like Mephistopheles, AI usage offers educators both power and convenience, but at what cost? 

A recurring theme in the use of AI in education is that it must be used critically and responsibly, especially given its impact on the student experience and its interactions with current pedagogy. In this use case, irresponsible AI use actually leads to inequities; why would we give some students access to the great literary canon, and others a robot-created summary? 

The bright side is that there is a path to redemption for those who want to use AI to effectively support student access to complex texts, and for what it's worth, Diffit does an incredible job of providing these resources. 

It may seem obvious, but AI tools can be used to create the scaffolds that provide access. Graphic organizers, note-takers, vocabulary reference sheets, and sentence starters are just the beginning. Here are some higher-leverage ways to use AI, in alignment with learning science, to help students access complex texts:

1. Have AI identify passages from the text that will trip kids up and use those passages as a basis for your syntactical and vocabulary instruction. 

2. Use AI to chunk, not change, the passages. Take it a step further by using “/” to separate phrases and create embedded fluency scaffolds that actually prove comprehension. 

3. Use AI to identify passages that could be suitable for fluency practice.

4. Not good at think-alouds? Have AI create your think-aloud script to model good reading of a complex text. 

5. Create word sets based on the text for deep morphological or connotative study

To avoid a Faustian tragedy, we must remember that AI is our Mephistopheles: a powerful servant, but a dangerous master. It should support thinking, creation, and access — not replace them.