The image of Sally the robot standing in a classroom is doing exactly what it was designed to do: capture our imagination. A humanoid machine teaching fractions to fifth graders feels like the future. It feels like innovation. It feels like a solution.
It isn't. And that's the real story hiding beneath the hype.
EdTech has always operated on a particular logic: identify a problem in education, build a technological intervention, deploy it in schools, and declare victory. The problem with teacher shortages? Deploy AI tutors. The problem with standardized instruction? Build adaptive learning platforms. The problem with educator burnout? Add more software to their workflows.
We are so committed to this technological-solution framework that we've stopped asking whether the underlying assumption is correct. We assume that if we can engineer our way around teachers, we should. We assume that more platforms, smarter algorithms, and lifelike robots represent progress.
Here's the uncomfortable truth: none of this addresses why teachers are leaving the profession in record numbers.
Recent commentary has rightfully pointed out that educators should have more agency in their tools and workflows. That vendors shouldn't dictate pedagogy. That platform success depends on actual instructional design, not just fancy features. These are good insights. They're also insufficient.
Because the real problem isn't which tool teachers use. It's that we've asked them to use too many.
Teachers today manage learning management systems, AI assessment tools, attendance software, parent communication apps, and now, increasingly, AI tutoring systems. Each tool promises efficiency. Each tool requires training, integration, and troubleshooting. Each tool collects data about students that educators may or may not control. The cumulative effect isn't liberation. It's cognitive overload.
We measure EdTech success by adoption rates and engagement metrics. We should measure it by whether it reduces the number of tabs a teacher has open during their planning period.
The structural shift happening beneath the robot headlines is this: EdTech has quietly shifted from supporting teaching to replacing teaching. Not in a science-fiction sense where robots teach everything. In a more insidious sense where the intellectual and relational core of teaching gets fragmented across disconnected systems, each one claiming to optimize a piece of the process.
Sally the robot might actually be good at drilling math facts. An adaptive platform might genuinely identify learning gaps. An AI system might even provide useful feedback on essays. But when you require teachers to orchestrate across all of them, while also managing actual human relationships with students, planning culturally responsive lessons, handling behavioral issues, and documenting everything for compliance, you've created an impossible situation.
We then blame teachers for not adopting innovations fast enough, or we blame the tools for being imperfectly integrated.
This is backward. The constraint isn't teacher resistance or platform limitations. It's cognitive bandwidth. It's the number of systems a human being can reasonably manage while doing the actual work of teaching.
Until EdTech vendors and school administrators address this structural reality, robots in classrooms will remain a distraction from the real problem: we've automated around teachers instead of automating for them.
The future of educational technology shouldn't be measured by how much it mimics human interaction. It should be measured by how much it actually reduces friction in the work that teachers find meaningful. That means fewer platforms, not more. That means deeper integration, not broader adoption. That means asking educators what they need to quit doing, not what new thing they can do with artificial intelligence.
Sally might be innovative. But she's not the innovation we need.