Leaving my school building the other day, I had an unexpected realization: Perhaps a computer was a more effective teacher than I currently was. The thought unnerved me, and still does as I’m writing this. I’m a nearly 13-year veteran educator dedicated to reflecting upon and refining my teaching craft. But I’m now considering the real possibility that, for at least part of a class period or school day, a computer could—and maybe should—replace me.

For the past several weeks, I’ve begun class with a simple routine: Students enter the room, grab a new Chromebook, log on to the Reading Plus program, and spend roughly 20 minutes working at their own pace. I stroll around the room and help with technology troubleshooting or conference with students, quietly chatting about academic progress or missing work. I’ve also found myself pausing, marveling at what this program promises to accomplish: meeting students where they are academically and, at least in theory, helping a wildly diverse group of students improve their literacy skills.

Developments in education technology promise to assist teachers and school systems in supporting struggling students by providing individualized instruction. But at what cost? As a teacher, it’s difficult to adapt to and embrace a machine that—at least for part of the time—takes over for me. The processes of teaching and learning are complex and innately human; I value the time I take to develop relationships with my students. But it’s hard not to wonder if that time could better be spent with adaptive learning technology.

My third-period sophomore English class at Fern Creek High School in Louisville, Kentucky, contains a wonderful mix of students hailing from the neighborhood and around the globe—my students represent Jordan, Afghanistan, Democratic Republic of Congo, Tanzania, Russia, and Mexico. I’ve thoroughly enjoyed getting to know how students arrived in our classroom in addition to hearing about their hopes, fears, and dreams. With this diversity also comes a huge range of student ability. Computerized reading assessments and other benchmarked tests reveal that roughly 90 percent of my class is behind grade level in reading.

About half of those students are at least four grade levels behind. My own anecdotal observations support this challenging reality as well. And across the country, only 34 percent of eighth-graders scored proficient or above in reading in 2015 according to the Nation’s Report Card. School districts’ attempts to improve literacy achievement are pervasive, and our school administration’s mandate to employ Reading Plus in most of our freshman and sophomore English classes reflects this.

I’d love to be able to provide individual instruction to my third-period class. One problem—and it’s a big one—is that I don’t know how to teach reading to students who are either new to the language or far behind grade level. And I know I’m hardly alone as a high-school English teacher in this tenuous position. I’ve earned an undergraduate degree in American literature, a master’s in teaching, and master’s in English literature. Yet these credentials haven’t equipped me with the necessary background or skills to significantly improve my students’ reading ability. I’m not trained as a reading specialist. Even if I were, how could I possibly create 27 customized lessons? Maybe Reading Plus can do some of what I can’t.

During the independent, silent work periods at the start of my class, the program adapts to students’ reading speed and comprehension ability, creating a customized scrolling illumination—imagine a rectangular flashlight beam only highlighting the text your eyes scan. Many students seem to embrace this moving target; at the least, they are more physically engaged with reading than ever before, and the program seems to be motivating a clear majority of students.

Reading Plus is emblematic of a growing trend toward personalized learning in public education; it’s the idea that schools can better serve students by providing more customized instruction. The term personalized learning refers to a vast array of approaches to education; examples include a high school in Deer Isle, Maine, and its radical curriculum overhaul to meet needs of individual learners in more creative ways, as well as San Diego’s High Tech High, where student-designed, long-term passion projects are paramount to the learning process.

Personalized learning, however, often manifests itself in school districts in less dynamic ways than in Maine and at High Tech High. The initiatives often become software or technology-based, with digital “instruction” adjusting based on competency levels or skills of its student users. It’s not about student passion or authentic projects—it’s all about remediating and measuring specific academic skills.

And as I’ve experienced first-hand, the role of teachers shifts dramatically with the adoption of these adaptive programs. Instead of a teacher striving to know a student on multiple levels—from understanding the nuances of his or her academic skills, to building positive relationships and crafting learning experiences based on more than numerical reading scores—educators are on the sidelines while a machine takes over. Personalized learning often becomes inherently impersonal; it’s a sterile approach to messy, complex classroom processes. And there’s also big money at stake for education-technology companies and curriculum publishers who are taking advantage of pressure to increase academic achievement.

According to this 2014 Education Week report, the Federal Department of Education Race to the Top competition awarded 16 school districts $350 million dollars to support efforts to personalize learning, often including adaptive software and digital tools as part of their plans.

For example, Miami-Dade Public Schools’ plan included buying access to Carnegie Learning’s Mathia, a program that “tutors” middle-school students in math. Carson City, Nevada’s, school system included a plan to incorporate MasteryConnect, which, according to report, is updated in real time as students take assessments, looking at mastery of learning targets (or specific academic skills). I wonder if educators in these locales are feeling as conflicted as I am.

Critics of the software-driven personalized-learning trend, including the author Alfie Kohn and FairTest, an organization dedicated to curtailing misuses and flaws of standardized testing, contend that there are significant problems with this approach. Kohn laments school districts’ focus on improving test scores as a catalyst in software adoption. One of the issues addressed in this FairTest post is that “frequent online student assessments require teachers to review copious amounts of data instead of teaching, observing and relating to students.” I agree with both of these criticisms, particularly the idea of losing more opportunities for human interaction in favor of customized screen time.

In 2014, I wrote a piece for The Atlantic titled “My Students Don’t Know How to Have a Conversation,” arguing that students’ reliance on screen time is detracting from their ability to communicate verbally. And now school systems are adopting programs designed to keep students glued to yet another screen for reading practice, which, by design, is a closed system. With Reading Plus, students do not have the shared experience and discussions after reading the same text, like when we analyze Kurt Vonnegut’s short story “Harrison Bergeron” or The Color Purple together. It’s all individualized, silent work. While we are still a community of learners, it feels less dynamic, even if students are making incremental reading gains according to the program.

For struggling readers and writers, it’s understandable that teachers, schools, and systems are striving to do whatever it takes to improve literacy levels. But whether struggling students are better off graduating from high school having been remediated by personalized-learning software versus more dynamic learning experiences, even if their reading skills marginally improve, remains an open question. I’m hopeful that this blended approach to teaching and learning—the combination of using technology-assisted activity and more traditional face-to-face methods—will be useful for my students. And I wasn’t always open to this possibility.

When I first read Michael Godsey’s essay for The Atlantic, “The Deconstruction of the K-12 Teacher,” a few years ago, I scoffed at the idea of teachers being replaced by classroom technology facilitators. Godsey writes, “The ‘virtual class’ will be introduced, guided, and curated by one of the country’s best teachers (a.k.a. a ‘super-teacher’), and it will include professionally produced footage of current events, relevant excerpts from powerful TedTalks, interactive games students can play against other students nationwide, and a formal assessment that the computer will immediately score and record.”  

In Godsey’s vision, those who currently serve as classroom teachers—like myself—would be replaced or forced to make radical changes in becoming a facilitator instead. Yet in the world of software-driven personalized learning, Godsey’s “super-teacher” isn’t even needed—only folks who can keep students behaved and on-task. I’ve reread the piece and agree with some of it’s conclusions: There’s no doubt the role of teachers is changing rapidly in many school districts towards more facilitation. Like Godsey, I’d struggle to tell a young teacher in training what to expect in the coming years—but there’s no doubt that blended learning will only increase in popularity. For now, I’m okay with my changing role, and it’s too early to tell if Reading Plus is worth the time and students’ effort.  

As I write my lesson plans for next week, I chunk out the daily time needed for students to engage with their personalized learning. I tell myself I’m still needed for the 45 minutes they aren’t tracking the illuminated scrolling target. I can still do my best to impart a love of writing, attempt to spark passions, encourage curiosity, foster discussions, smile, laugh, and interact with the students in ways a screen can’t, even if Reading Plus “knows” more technical information about their reading levels than I ever could.