What we lose when AI replaces teachers

Key points:

A colleague of ours recently attended an AI training where the opening slide featured a list of all the ways AI can revolutionize our classrooms. Grading was listed at the top. Sure, AI can grade papers in mere seconds, but should it?

As one of our students, Jane, stated: “It has a rubric and can quantify it. It has benchmarks. But that is not what actually goes into writing.” Our students recognize that AI cannot replace the empathy and deep understanding that recognizes the growth, effort, and development of their voice. What concerns us most about grading our students’ written work with AI is the transformation of their audience from human to robot.

If we teach our students throughout their writing lives that what the grading robot says matters most, then we are teaching them that their audience doesn’t matter. As Wyatt, another student, put it: “If you can use AI to grade me, I can use AI to write.” NCTE, in its position statements for Generative AI, reminds us that writing is a human act, not a mechanical one. Reducing it to automated scores undermines its value and teaches students, like Wyatt and Jane, that the only time we write is for a grade. That is a future of teaching writing we hope to never see.

We need to pause when tech companies tout AI as the grader of student writing. This isn’t a question of capability. AI can score essays. It can be calibrated to rubrics. It can, as Jane

said, provide students with encouragement and feedback specific to their developing skills. And we have no doubt it has the potential to make a teacher’s grading life easier. But just because we can outsource some educational functions to technology doesn’t mean we should.

It is bad enough how many students already see their teacher as their only audience. Or worse, when students are writing for teachers who see their written work strictly through the lens of a rubric, their audience is limited to the rubric. Even those options are better than writing for a bot. Instead, let’s question how often our students write to a broader audience of their peers, parents, community, or a panel of judges for a writing contest. We need to reengage with writing as a process and implement AI as a guide or aide rather than a judge with the last word on an essay score.

Our best foot forward is to put AI in its place. The use of AI in the writing process is better served in the developing stages of writing. AI is excellent as a guide for brainstorming. It can help in a variety of ways when a student is struggling and looking for five alternatives to their current ending or an idea for a metaphor. And if you or your students like AI’s grading feature, they can paste their work into a bot for feedback prior to handing it in as a final draft.

We need to recognize that there are grave consequences if we let a bot do all the grading. As teachers, we should recognize bot grading for what it is: automated education. We can and should leave the promises of hundreds of essays graded in an hour for the standardized test providers. Our classrooms are alive with people who have stories to tell, arguments to make, and research to conduct. We see our students beyond the raw data of their work. We recognize that the poem our student has written for their sick grandparent might be a little flawed, but it matters a whole lot to the person writing it and to the person they are writing it for. We see the excitement or determination in our students’ eyes when they’ve chosen a research topic that is important to them. They want their cause to be known and understood by others, not processed and graded by a bot.

The adoption of AI into education should be conducted with caution. Many educators are experimenting with using AI tools in thoughtful and student-centered ways. In a recent article, David Cutler describes his experience using an AI-assisted platform to provide feedback on his students’ essays. While Cutler found the tool surprisingly accurate and helpful, the true value lies in the feedback being used as part of the revision process. As this article reinforces, the role of a teacher is not just to grade, but to support and guide learning. When used intentionally (and we emphasize, as in-process feedback) AI can enhance that learning, but the final word, and the relationship behind it, must still come from a human being.

When we hand over grading to AI, we risk handing over something much bigger–our students’ belief that their words matter and deserve an audience. Our students don’t write to impress a rubric, they write to be heard. And when we replace the reader with a robot, we risk teaching our students that their voices only matter to the machine. We need to let AI support the writing process, not define the product. Let it offer ideas, not deliver grades. When we use it at the right moments and for the right reasons, it can make us better teachers and help our students grow. But let’s never confuse efficiency with empathy. Or algorithms with understanding.