A few years ago, on a subway train, I watched a child cry because a paper picture book would not respond as she swiped her hand across the page. Smartphones and iPads now deliver content that is far more dynamic and stimulating, and the old picture book no longer satisfies a child's curiosity. Before we have even adjusted to this digital shift, artificial intelligence has arrived at the scale of a transformation, posing a still larger question. What should we teach our children, and how? As AI seeps rapidly into daily life, the worries of parents and educators are growing. What happens if children hand off all their homework to AI, or talk with AI instead of with people? The loss of the ability to think for oneself is another concern.
A recent survey by MIT Technology Review, the science and technology magazine of the Massachusetts Institute of Technology, asked young people aged 10 to 18 what they thought of AI, and the findings are striking. They view AI far more realistically, and in more layered ways, than adults expect. Of those surveyed, 57% said they used chatbots to look up information, 54% for schoolwork and 47% for fun and entertainment. Only 12% said they used them to seek emotional advice. They resisted the idea of AI standing in for their own thinking and creativity. One student said she used AI to correct the grammar of her own writing and to find and fix bugs in game code she had written herself, but did not let AI create the story itself, because she wanted her creativity to remain her own. It is reassuring that young people see AI as a tool that offers technical help, not something that should take over their own creative work.
As generative AI systems such as ChatGPT, Claude and Gemini grow steadily more powerful, educators face harder questions. At first they debated how much AI to permit in coursework and research. Now they are asking how to teach students to use it well. The possibilities AI brings to education are enormous. Feed in lecture videos, textbooks and reference materials, and it can generate exam questions and practice problems, and it is possible to build systems that explain, as a teacher would, whatever a student has failed to grasp. Used well, AI can extend to every student the kind of educational opportunity that once required money and the right address — expensive private tutoring, top-tier cram schools.
As many educators already agree, what matters in education in the age of AI is not how much knowledge one can remember. What matters is the ability to frame good questions, to verify and criticize the answers AI produces, to judge which of several answers is right, and to take responsibility for that judgment. And the process of thinking itself must never be handed over to AI. Between a student who asks AI a question and immediately writes down the answer, and a student who first thinks the problem through, then debates it with AI, asks again, compares it against other AI material and reaches the final conclusion alone, the gap will grow enormous over time. The first becomes a person dependent on AI; the second becomes a person who uses AI as an amplifier of ability. From now on, assessment must weigh not only the result but the thinking that led to it. The whole process will be assessed: whether AI was used and what questions were asked, which answers were taken or discarded and why, what was found to be wrong, how the student's own thinking changed and what conclusion was reached.
In the end, the goal of education in the age of AI must not be to produce students who are skilled at AI. Even with AI at hand, they must be able to think for themselves and to doubt the answers AI gives. They must ask better questions than AI does and interact with it in ways that draw out better answers. Most important of all, we must raise students who can judge what to leave to AI and what to keep as the work of human beings. We must teach them to use AI as the most powerful of tools while remaining, to the end, the owners of their own thinking and judgment. Facing an era of AI transformation, it is time to think seriously about how our current systems for early childhood and primary and secondary education, and our college admissions system, must change.
