Teaching With Humanity: Educators Must Offer What AI Cannot

The question arrives quietly, almost as an afterthought: What does it mean to teach when you're deliberately not using artificial intelligence? The answer, it turns out, is nothing like what came before.

We live in a moment of genuine ambivalence about AI in education. Some educators worry about the environmental cost of training these systems, others about the corporations controlling them, still others about gaps in our understanding of what these tools actually do to how students think. The reasonable impulse is to pump the brakes. But even as institutions debate whether to embrace or restrict AI, something more urgent is happening in classrooms: students are already arriving with transformed expectations. They've grown accustomed to instant answers, to outsourcing cognitive labor, to the promise that struggle can be eliminated. Teaching without AI, then, can't simply mean ignoring these developments and hoping students forget what they've learned about efficiency. It has to mean something deliberately different.

The trap is thinking that teaching without AI means returning to what worked before. It probably didn't work well enough, or we wouldn't be in this position now. If the old approaches had made a convincing case for the value of human struggle, discovery, and growth—for the messy, winding path to understanding—we might not now be drowning in arguments about efficiency, misinformation, and the seductive offer to remove meaningful work from our hands. The task isn't restoration. It's reinvention.

One place to start is with personalization, which AI vendors promise but which teachers can actually deliver. The technology claims it can tailor learning to each student's individual needs. That's worth doing—just not through algorithms. It means asking: How much agency can we give students to identify their own learning goals? How can we design course projects around what matters to them personally? How responsive can we be to their individual struggles and progress, creating genuinely different learning arcs for different people? This requires knowing your students—surveying them, talking with them in class and in office hours—and then building flexibility into how you teach.

The practical work happens in assignment design. Instead of a handful of high-stakes exams or papers, offer many smaller opportunities to experiment, fail, adjust, and try again. Break major projects into discrete parts. Allow students to revise work and resubmit. This isn't grade inflation; it's creating space for the actual process of learning, which is rarely linear. It's also worth examining your grading system honestly. To what extent does it reward outcomes alone, and to what extent does it value the labor of getting there? Process grading and completion grading aren't soft options—they're statements about what you believe learning actually is.

Reflection matters too, and it's something humans do better than machines. Model it yourself: think aloud about your own work, your own discoveries, your own dead ends. Invite students to reflect informally on their learning through prompts in assignments. Meet with them one-on-one when you can, so reflection becomes conversation. Guide them toward specific insights: the connection between discomfort and breakthrough, the irreplaceable value of doing difficult work yourself at each stage, the difference between understanding that arrives easily and understanding that's been earned. These aren't abstract lessons. They're the texture of what it means to learn.

But there's a limit to what individual teachers can fix. Students are carrying institutional weight—obligations piling up from multiple courses, multiple demands, multiple systems all competing for their time and attention. The broader question higher education needs to ask itself is structural: What are we actually asking of our students, and does it leave room for expansive learning? Do our expectations create space for struggle, or do they demand efficiency? Teaching without AI can't just be about individual teachers saying no to a technology. It has to be about institutions and classrooms deliberately choosing to make space for what AI can't provide: the experience of reaching understanding through your own effort, the pleasure of discovery, the growth that comes from wrestling with difficulty. That's teaching with humanity, and it's the only kind worth defending.