Are we raising kids who can't think without an AI prompt?

The headline asks a question that feels like a moral panic, but the real stakes lie in the classrooms of Lagos, Nairobi, and Accra where a teacher‑shortage crisis has turned learning into a survival game. When a chatbot pops up with a ready‑made answer, the child no longer wrestles with a problem; the struggle that builds critical thinking evaporates. Not whether AI will ever replace the teacher, but whether it will replace the process of learning itself.

On September 3, New York City, the world’s largest public school system, outlawed AI use for grades up to eight. The move sparked a flurry of commentary across the Atlantic. In a recent episode of Delve Into AI, ADONIJAH NDEGE explains that the ban would be a mistake for Africa, where the alternative to an AI tutor is often no tutor at all. The policy, intended to protect young minds from cognitive shortcuts, ignores the reality that many African schools have no qualified teachers to fill the void. The ban would therefore deepen an existing inequity rather than solve it.

NDEGE’s conversation with Kurt Gammelgaard Nielsen, chief executive of EdTech Denmark, reveals a deeper philosophical divide. Nielsen argues that AI should augment, not replace, human learning. “Children should be empowered to think critically, explain deeply, and learn more effectively,” he says. Yet the very design of chatbotsinstant answers, no scaffoldingpromotes cognitive offloading. The child learns to accept a finished answer without questioning its validity or exploring alternative solutions. In the long run, this erodes the very skills that education is meant to cultivate.

NDEGE proposes a counter‑intuitive solution: let the machine ask the questions while the student generates the answers. This flips the traditional AI‑teacher dynamic on its head. Instead of a chatbot handing out solutions, it would present a problem, prompting the learner to apply reasoning, research, and creativity. In practice, this could be as simple as a classroom app that generates a math puzzle or a history debate prompt, then records the student’s response for later reflection. The machine becomes a question‑generator, not an answer‑giver, preserving the struggle that is essential to learning.

I predict that by 2028, African EdTech firms will pivot from “AI tutor” models to “AI facilitator” platforms. Companies that can embed question‑generation algorithms into low‑bandwidth devices will capture the market, especially in rural areas where teacher shortages are acute. This shift will also satisfy regulators who fear that AI could undermine critical thinking, giving governments a way to adopt technology without compromising educational integrity.

The debate over AI in classrooms is not a binary choice between technology and tradition; it is a question of how we use technology to shape minds. If we let AI hand out answers, we risk creating a generation that can recite facts but cannot interrogate them. If we let AI ask the questions, we harness its power to amplify human curiosity and resilience. The choice is clear: we must design AI tools that ask, not answer, and in doing so, we give African children the cognitive tools they need to thrive in a digital world.