The craze for artificial intelligence is sweeping into the homework of primary and secondary school students. A recent media survey found that some students have begun using generative AI (AIGC) products on the market to help them write essays and reading reflections, and some have even developed the habit of "turning to AI whenever there's a problem." Such "lazy" scenarios inevitably bring to mind the previously widespread "photo-search homework apps," which is truly worrying.

Under the demonstration effect of ChatGPT, various AIGC products have sprung up like mushrooms after rain. For many primary and secondary school students, AIGC products are enormously attractive. On the one hand, they are young and curious, with a higher acceptance of novel things; on the other hand, supported by large-model technology, AIGC products can serve as tools for compressing and retrieving various kinds of knowledge and information, and objectively can indeed take on some of the functions of teachers. But technology is a double-edged sword: while we enjoy its convenience, we must also be alert to its hidden risks. Whether artificial intelligence or homework-search tools, they can indeed resolve doubts and fill gaps, but young people have weak self-control, and when faced with tools that deliver answers "the moment you speak" or "the moment you snap a photo," they can easily develop an over-reliance syndrome. Usually it goes unnoticed, but the moment an exam comes, the truth is laid bare.

The development of artificial intelligence technology is an irresistible trend, and empowering education with it holds infinite possibilities. Judging from existing technology products, the images and videos generated by ChatGPT and Sora can already carry rich information and value; appropriately applied in the field of education, they can effectively train students' abstract thinking and imagination. At the same time, large models that are like "encyclopedias" can expand the breadth of knowledge and accelerate the dissemination of information, and once popularized will greatly benefit the promotion of educational equity. Given the various characteristics of AI technology, a "one-size-fits-all" ban on student use is neither realistic nor reasonable. In this regard, perhaps we should refer to the management approach of short-video and livestreaming platforms and introduce a "youth mode" or "minor mode," "limiting" the number of uses and the duration per day, to prevent students from becoming addicted over the long term and to avoid dependence and inertia.

From online classrooms to homework-search apps and now to AIGC, along with each round of technological revolution, ever newer and smarter applications will continuously challenge students' traditional ways of learning. Whether the effect of education embracing new technology is positive or negative depends crucially on the awareness of those who use it. From this perspective, staying clear-headed and self-disciplined in the digital age, possessing information and media literacy, and learning to use digital tools reasonably and in moderation will become a basic survival skill for citizens and a required course in children's growth. Therefore, teaching students how to use artificial intelligence reasonably, lawfully, and with restraint is an important task facing the education sector today.

"The value of education lies not in learning many facts, but in training the brain to think." In promoting the development of "AI + education," progress in technology itself is of course important, but helping primary and secondary school students better learn and adapt to constantly changing new technologies is where the task of education truly lies.