Overview:
We are rushing towards AI in education without seriously assessing the risks, and looking at the warning precedents (social media, digital classrooms), we must slow down before real generational damage is done.
I see a lot of people on LinkedIn (and elsewhere) writing about how great AI can be, personalized learning, making learning fun, enabling conversation, discovery, and a host of other potential benefits. Other benefits are available. And I agree with most of the posters in principle, as a certified paying user of ChatGPT, an AI writing teacher, and a parent. The biggest thing missing from the hype is balance.
What if we get AI from the wrong education? Why should we slow down? When will we know the problems? How will errors appear? Who will be affected? What else is there? Captions appear now, with similar writing and design features.
Are these the right questions? When AI in education goes wrong, it is too late, and there is damage done to the education and development of young minds. A generation of clones. Prevention is better than cure. Or we should be asking, “How do we prevent AI from going bad in education?”
Much has been written about Alpha School and others using AI to engage and support students. But these are small examples with parents involved (or at least parents who can pay for support). Scaling is not mentioned. I asked in several LinkedIn posts, without much response. Charter schools had promise, too, but without support, many high school graduates have languished in university (where they are controlled by factors such as choice, home, etc.). It’s a mixed bag, similar to what’s happening with AI schools emerging now.
Examples are given of how to make homework fun and how facilitators are using AI with their children. Of course, AI can make homework fun, engaging, and educational. However, not many parents are actively involved in their child’s education. Out of 16 children in my son’s 4th grade, only two of us are involved in our children’s learning. Even then, my son says I do too much and just writing simple information is fine because that’s what his friends do! Among my university classes, less than 1 in 5 students report parental involvement in their education. As the old saying goes, friends are more influential than parents.
We have seen with digital classrooms that learning can be accessible but not enhanced, or limit the internalization of information. There is an appearance of knowing without actually knowing. A 2024 study showed that students who use AI learning assistants do worse on tests. No wonder Sweden limited its digital classroom and brought back a lot of book time.
Social media, too, had important promises, but it has caused serious psychological and even physical problems for young people. The Australian belt and limitations of other countries may seem extreme to those of us who grew up before 2000, but those generations (X, Y, and some boomers) know how to handle technology much better. They experienced its evolution. A teenager in 2025 has had the internet and social media their entire life. The only change is the platforms. AI seems new, but it is a continuation (or part) of machine learning.
Therefore, some of the more mature supporters can use AI for great benefits. Students with disabilities or language needs can also benefit greatly from AI. But there are many left behind. Those who use LLMs to find specific answers rather than as a tool for thinking, or thinking about answers. It is not a technology per se, however, to rely on a short line. Usage is important.
Success in education is measured across generations at the societal and individual levels. We see the gentle parenting of the 1990s reflected in the ultra-gentle parenting of the 2020s. Children as accessories. Combined with the proliferation of technology that can act as babysitters (from multi-channel TVs to YouTube-enabled tablets), parents are choosing not to spend their time reading to children. Declining reading rates have accelerated over the past 10 years because those 1990s kids are now becoming parents with more toys to direct attention and responsibilities.
The rise of block reading in the 1980s – 1990s has led to major policy problems now. Semesters are common, and courses are divided into different parts. Spaces are ignored. As a result, seemingly intelligent (university-educated) people falter in simple concepts. For example, climate is long-term (typically 30 plus years), and climate is short-term (typically less than a year). This concept of high school is missed by many. So, when a PhD expert says climate change is a hoax, it misleads the undergrads. The so-called expert does not have a complete understanding but can influence bad policies on the same people. The reign of chaos, even among intellectuals, began many years ago, and has only gotten worse. And it could be worse as early studies show that AI is rewiring our brains!
Those who challenged the changes were, as always, called the old-timers. Holding back progress to preserve their ropes. There is, like all things, some truth in these statements. In all things there is good and bad. Balance is what is needed. The Luddites were not against technology, but implementation.
However, after more than 150 years of public education, shouldn’t we, the teachers, think long term and use the great tools at our disposal to conduct balanced evaluations? A recent OECD report said Edtech has not delivered on its promises (especially to the masses) after 10 years of ubiquitous implementation. Basic literacy and numeracy skills are required to master AI tools. What’s that about putting the cart before the horse?
While I use AI, I teach my students to use AI, and I engage my children with AI, I am also a pragmatist. Parents, schools, and the public need to ask questions about if we’re getting AI in education wrong. There are many recent examples of making things worse. There is no need to recall eugenics and mental health studies conducted 100 years ago (although they are instructive). Smoking was once promoted as healthy and ignored for generations. Walking barefoot in the grass with animals was once considered backward and unhealthy, but is now known to be physically and mentally beneficial. A list of examples of misplaced hopes exists.
So, with just over three and a half years of ChatGPT and the widespread use of AI, we should take the time to think and give concrete answers about the direct and generational impact of getting AI wrong in education.
