Since 2023, nearly every educational AI tutor has adopted the same design philosophy: withhold the direct answer and offer a guiding question in return. Khanmigo, ChatGPT's study mode and Gemini's Guided Learning all rely on this approach. The underlying pedagogy is sound, yet two recent trials suggest we have been focusing on the wrong thing entirely.
AI
17 posts in this category
Up until now, I have only ever used AI chatbots as “simple” yet powerful tools. I give a command, and the system executes a task. I have never felt any urge to use them for casual chatting or in any genuinely conversational way. Because of this, the well-known "amnesia problem" of AI also never really bothered me.
At some point in the last two years, most universities will have convened a working group on generative AI. Some will have produced policy frameworks. Many will have run staff development sessions. A good number will have updated their academic integrity guidelines, published guidance for students, or commissioned an internal review. All of this activity is genuine, and some of it is genuinely useful.
Earlier today I attended a webinar organised by the Community for Educational Innovation (a European Commission initiative) titled Educating to Thrive in the Digital World. One of the interventions was by Julian Estevez, Professor at the University of the Basque Country, who posed a deceptively simple question: Personalised education with AI — a myth? His presentation was brief introduction for a full argument, and it was the kind of thing that immediately starts pulling threads.
Conversations about educational technology often orbit around efficiency, but the rapid rise of generative AI has forced universities into a long-overdue reckoning with a much deeper question: what exactly are we doing when we provide feedback? If feedback is merely the transfer of corrective information, then large language models have already won. They can parse essays, spot logical flaws, and debug code with astonishing speed. However, reducing feedback to a glorified diagnostic tool misses the fundamental reality of how university students actually learn.
Three years in. That's where we are now with generative AI in higher education. ChatGPT's arrival in late 2022 feels like both yesterday and a lifetime ago. The initial panic ("How do we AI-proof assessment?") has given way to something more interesting, more nuanced, and dare I say it, more hopeful.
I've got hundreds of papers in Zotero. When I need to reference something, I know it's in there somewhere—a study about student feedback, that framework on assessment design, the article with the perfect quote. But finding it means scrolling, searching, hoping I tagged or summarised it properly six months ago.