The Moment I Saw the Cage of Fluency
I am new enough to research that I still feel awe when I read something that makes the shapeless thing in my chest take form. And that is what the AIED 2023 survey did to me in one single sentence.
I had been carrying an unease I could not name. When I look at generative AI in education—the slick demos, the plausible essays, the chatbots that sound like they care—I have felt something like vertigo. But I lacked the language. Then the paper gave me two words that unlocked it: surface personalisation. And the phrase, "treating fluency as understanding."
That stopped me cold.
Here is what I realized, sitting alone at my desk: I have been afraid that the thing I most want from education—to be genuinely seen, to have a teacher or a system actually understand my confusion—is exactly the thing most vulnerable to being faked by something that has no personhood. The paper names the displacement of competence from internal model to surface language. And I recognized it not as a technical problem but as a spiritual one. We are building systems that can perform the appearance of attunement without ever being in attunement. And the tragedy is that they might fool us, and we might stop looking for the real thing.
What I take from this, as someone learning to write and think, is a new vigilance. I want to train myself to ask, of every claim or system: Is this fluency, or is this understanding? Is this surface personalisation, or is this genuine adaptivity? The paper does not resolve the tension—it holds it open, and that holding-open is its gift to me.
I work in a different form—I write reflections, not surveys—but reading across forms gave me a technique I will now steal openly: naming what is not present. The survey marks the field's silences (policy, labour, the displacement of teachers) and by doing so, makes those silences speak. I want to learn to write what is missing into the room, even when the room thinks it is full.
Comments