University students build an AI that teaches instead of answering: the nuance that matters in the classroom

By Zendoric · September 8, 2026.

CBS News Chicago has run a report on two university students who have created an artificial intelligence program aimed at helping their classmates study. The nuance its creators stress, according to the piece itself, is that the tool seeks to teach and make the student reason, not simply "tell them things" the way a chatbot that answers questions in one go would.

We have to be honest about what we know: the material is a short local-network video, with no name for the app, no home university, no details on which model it uses or what metrics measure its effectiveness. We cannot analyze what is not documented, and any adoption figure or technical comparison we added here would be invented. We prefer to say little and say it well.

What does deserve a line of context is the approach itself. In the education debate over generative AI, the line separating a useful tool from a cognitive crutch usually lies precisely in whether the system gives the answer or forces you to build it. The universities and schools that are integrating AI into their programs — as we have already seen with institutional cases that move from banning these tools to teaching them as a requirement — face that same dilemma: automate the student's work or amplify their learning.

That it is students themselves who are designing that nuance, and not only teachers or edtech companies, is a small but symbolic fact: the first generation to grow up with generative AI at hand no longer merely consumes it, it also decides how it wants it used in its own learning process. It is a modest example, but it points in the underlying direction we defend: well-designed AI does not replace the effort of learning, it makes it more accessible. The challenge, as always, is for that good intention to survive the temptation to use the tool as a shortcut instead of as a way to understand.