by Byron Scaf
Speaking with a respected science teacher last week, she vividly recalled her own student days during a parent-teacher interview when her teacher remarked that she could do “so much better” than become a teacher. Her parents were mortified when she enrolled in her Master of Teaching, but luckily for her students, that didn’t hold her back.
Our collective dismissiveness of the teaching profession borders on a national sport. It permeates culture, politics, and business. The most insidious examples come from within. For years, the “EdTech” community—those creating software-powered services for schools—have tried to work around teachers. This is evident in the products they create and their private conversations behind closed doors. The latest manifestation of this is the slap-dash use of Artificial Intelligence, which is ultimately doing all of us a disservice.
Silicon Valley is notorious for its “we know best” attitude. If you’ve been anywhere near education over the last two decades—as a student, parent, or teacher—you’ve likely witnessed the parade of uninspiring education software that, in one way or another, has been designed to replace teachers. “Don’t worry, the software will figure out where your students need help and then algorithmically serve them the next piece of learning.”
Of course, this is positioned as a tool to “help” struggling teachers. I might believe that more if, behind closed doors, executives at these companies didn’t disparage those same teachers. Now, we find ourselves in the midst of AI mania. At education trade shows, both here and abroad, companies are peddling their AI-powered replacements for teachers. Not that they’d ever call it that, of course, but in many cases, the implication is clear. However, an overreliance on AI to instruct is problematic. Teachers possess the historical knowledge and nuance that comes only from time in the classroom—qualities that AI simply cannot replicate.
Consider, for example, an AI providing direct, unmoderated, human-like feedback to students based on their responses to questions in an online platform. Even if AI could reliably offer correct feedback all the time, which it can’t yet, the whole concept is flawed.
Effective feedback is built on a trusting relationship. Students grow from feedback when they trust the interests and intentions of the person providing it. This relationship gives teachers essential context: Is the student having a rough week? Is there something going on at home? Is now the right time to push or hold back? Is this student ready to receive constructive feedback?
Pausing for a moment, it’s clear how absurd it is to think that all that’s required for effective feedback can be deduced from a student’s answers to a series of curriculum-aligned questions online. The last thing students need, and the last thing that will engage them, is a flood of AI-generated, low-quality feedback.
Instead, AI technology in education should serve as an assistive tool—one that always positions the teacher as the expert and keeps them in the loop with their students’ learning. Likewise, students should become skilled AI users, but never at the expense of genuine knowledge. As AI becomes more common, knowing how to use it effectively will be a crucial skill. However, like any tool, AI cannot replace the foundational knowledge students need now and in the future to succeed.
Students are complex, real-life human beings. They have good weeks and bad. Sometimes, they need a pep talk; other times, they need the hard truth. Sometimes, they’re ready for feedback, and sometimes, they’re not. Often, they just need someone to listen. Teachers intrinsically understand this and adjust to meet their students’ needs.
If AI is to fulfill its promise in education, we must approach it with the mindset of creating the ultimate teacher’s assistant. Let’s build that—not a clumsy replacement for teachers.
Let’s do it because teachers deserve “so much better.”
Byron Scaf is the CEO and co-founder of Stile Education.