I work with a lot of educators. I hear the same thing on repeat—“Our students are using AI, and they’re not even hiding it.” Some are worried. Some are curious. Most are asking the right question: What now?
Here’s where my mind keeps landing.
When the answer is always one click away, the real skill becomes knowing which questions matter.
The illusion of understanding
Back in the day, students had to hunt for information. They sat with the unknown a little longer. Today, ChatGPT gives you a decent answer before you finish the sentence. And that creates a subtle danger: students feel like they understand something before they’ve actually thought it through.
A recent MIT Media Lab study found that students using generative AI to write essays engaged in less cognitive processing, showed lower retention, and didn’t perform as well when asked to apply those ideas later.¹
It’s not that they got the facts wrong. It’s that they didn’t wrestle with them.
Safety brings curiosity back
There’s another side to this. And it’s hopeful.
In a March 2025 survey from Filtered, high school students said they loved asking “dumb” questions to AI because it removed the fear of judgment.² That matters more than we think. When students feel safe, they get curious again.
For teachers, this is the shift: using AI not to give the answer, but to make it okay to ask the question in the first place.
A new kind of learning
This spring, researchers studying 14- and 15-year-olds working on science tasks with ChatGPT noticed something surprising: it wasn’t the answers that made the biggest difference. It was the quality of questions the students asked.³ Those who used AI to explore and refine their thinking—not just grab conclusions—performed better and stayed engaged longer.
That’s the future. Not “how do I finish this worksheet?” but “how do I test an idea, challenge a claim, or tell if this source is credible?”
Teaching in an answer-rich world
Teachers will need to model a new kind of literacy—one that values discernment over recall. Less about getting it right. More about slowing down and staying with a question long enough to learn something real.
This doesn’t require a total overhaul, but it starts with things like:
Designing prompts that invite ambiguity
Giving credit for thoughtful revision, not just output
Helping students reflect on how they used AI, not just what it produced
Generative AI won’t destroy wonder. But it will change how we reach it.
In this new era, fast answers are everywhere. What we need—what our students need—are spaces that reward depth, pause, and asking the second or third question.
Because curiosity isn’t a given. It’s a discipline. And now more than ever, it needs practice.
Sources
MIT Media Lab, “Cognitive Load in Generative-AI-Assisted Essay Writing,” March 2025.
Filtered AI Education Survey, Student Use of ChatGPT in Secondary Classrooms, March 2025.
ArXiv preprint, “Generative AI and Inquiry-Based Science Education in Early Adolescents,” May 2025.
Bob Hutchins’s article offers a thoughtful, hopeful take on how AI can reignite our sense of wonder. His use of McLuhan’s tetrad provides a smart framework for understanding AI’s impact. The piece balances excitement and caution with clarity and grace.