You’d have to be living under a rock to have missed the sharp uptick in anti-edtech headlines over the last few months.
In March, Fortune led with “America’s math and reading scores tanked after schools ditched textbooks for screens—and AI could worsen the brain rot.” Earlier this month, Chalkbeat titled a story “The AI rebellion grows in NYC: Over 100 New Yorkers demand moratorium on AI use in schools at marathon board meeting.” This week, NBC News’ Tyler Kingkade published “The revolt against i-Ready,” a tour of parent, teacher, and student backlash against an assessment and instruction platform that is used in nearly a third of K-12 classrooms in the country.
The tone is reaching a fever pitch.
Altair Maine, an LA high school teacher, claimed tech was turning him into a “glorified babysitter” of “kids staring at Chromebooks.” Ward Wooden, a 14-year-old in Los Angeles: “I’m losing brain cells every time I do a lesson.” Ward’s father, John Allen Wooden describes what he views as a tech “stranglehold on U.S. classrooms.”
Although the headlines paint a binary picture, read more of the underlying interviews and a different story emerges. Neuroscientist Jared Cooney Horvath, in the same Fortune piece, added: “This is not a debate about rejecting technology. It is a question of aligning educational tools with how human learning actually works.” And Julia Nasef, the student member of the NYC panel, told her colleagues that many of her peers find AI helpful when used “intentionally”—urging the adults in the room to “support clear, student-centered guidelines for AI implementation.”
Headlines notwithstanding, this is the work of serious journalists and researchers trying to get it right. But the headlines are emblematic of an assumption that has dogged the edtech conversation for decades: that technology in school is one “thing,” and that we should decide, all-or-nothing, whether it belongs. In reality, tech in the classroom serves a multiplicity of purposes.
Although using tech to deliver content is arguably the least interesting application, it is often the highest profile. Tech can also enable high-volume, low-stakes repetition to build fluency with, say, “math facts,” vocabulary, or other rote skills (that students sometimes find maddening). Amidst the tumult, diagnostic and other attributes are often lost in the mix.
In some cases, adaptive (and increasingly AI-enabled) tools can help to surface patterns that a teacher with 32 students and a six-hour day might not be able to see on their own. Teletherapy can connect learners with speech language pathologists or mental health professionals and increasingly advanced speech recognition can actually “listen” as students read aloud, measure fluency or help to identify dyslexia or emergent reading difficulties.
As I have written in the past, the edtech industry is often its own worst enemy, but there’s a risk that with the sort of backlash Kingkade chronicles, the most promising innovations become the proverbial babies in the bathwater.
One promising structural innovation comes in the form of outcomes-based contracting (OBC). It’s not a new concept, but is gaining traction thanks to the work of the Center for Outcomes Based Contracting, which advocates for tying a share of vendor payment to whether the tool actually delivers under defined conditions—dosage, population, baseline data, target outcomes, all spelled out up front.
A March report from Digital Promise studying the Center’s first edtech cohort found that districts using OBCs saw students hit prescribed dosage at rates roughly 14 times the industry baseline, with one district reaching 95% under an Amira Learning contract. The point isn’t the contract mechanics. It’s what the mechanics force: both sides specifying the job the tool is being hired to do, the conditions under which it can plausibly do it, and the evidence that will determine whether it did. That is the discipline the field has been missing.
Which brings me to an often-told story in the world of tech.
In 1997, Garry Kasparov famously lost to IBM’s Deep Blue and, instead of retreating, invented “Advanced Chess“: human-plus-machine pairs competing against other pairs. The result was the finding that amateurs with modest laptops and a disciplined process consistently beat a grandmaster with a computer who lacked one.
The lesson is not that the machine wins. The lesson is that the interface between the human and the machine—the workflow, the judgment, the dosage—is where the value lives.
That is the conversation that edtech headlines have yet to convey. The question is not whether technology belongs in classrooms. The real questions are which uses, in what doses, with what professional learning, and under what diagnostic discipline.
We will not answer them by banning adaptive tools and AI, and we will not answer them by deploying it everywhere because a vendor said so. We will answer them the way Kasparov did: by building the workflow first, and letting the tool serve it.
This article is sourced from Whiteboard Notes, our weekly newsletter of the latest education policy and industry news read by thousands of education leaders, investors, grantmakers, and entrepreneurs. Subscribe here.
